26 Measurement and Testing Box and Whisker Plot Comparison Figure 9: Poor Precision
Precision of the study data was calculated in accordance with ASTM D6300, Standard Practice for Determination of Precision and Bias Data for Use in Test Methods for Petroleum Products and Lubricants, and is depicted in graphical format in Figures 10-13, with repeatability in Figures 10 and 11, and reproducibility in Figures 12 and 13. The precision of the study data is depicted in solid blue (HDXRF) or orange (ICP) lines. For context, the precision of ASTM test method D5708B (ICP after acid decomposition) has been added as a dotted orange line to all figures.
Figure 12: Nickel Reproducibility
Another way to compare both correlation and precision is by creating a box and whisker plot. This type of plot shows a quick graphical examination of sample results distribution. See Figure 14 for the anatomy of a box plot. The box encompasses the 25th through 75th percentiles, with a line drawn at the median. The larger the box the wider the sample distribution, therefore, a smaller box indicates tighter precision. If the 50% median bisects the center of the box, this indicates a normal distribution (bell curve). If it is closer to one end or the other, then the distribution is skewed. When comparing two methods, the relative size of the box is indicative of precision, and the median is used to infer correlation. In the example in Figure 14, method B has better precision than method A, and the methods show good correlation because their medians are similar.
Figure 14: Box Plot Example Figure 13: Vanadium Reproducibility
The results of this study show that Petra MAX, powered by HDXRF, is more efficient and delivers precision comparable or better than the ICP ASTM test method precision (D5708B) for nickel and vanadium in crude oil.
For both nickel and vanadium, the ICP repeatability and reproducibility results demonstrated poorer precision than expected, as shown in Figures 10, 11, 12, and 13. To understand why, there are a few possible explanations to consider:
Figure 10: Nickel Repeatability
ASTM D5708B method scope covers both crude oils and residual fuels. When the ASTM D5708B precision study was conducted, samples were included that covered not only crude oils, but also residual fuels. In our study, only crude oils were included. It may be that residual fuels are easier to prepare and analyze than crude oil by ASTM D5708B. The additional data from residual oil to the overall precision statement of ASTM D5708B may lower the overall precision from what you would see when only crude oil samples are included.
Sample D was a highly bituminous crude oil from Canada. This type of sample represents one of the most challenging sample types to measure by ASTM D5708B, because bituminous samples are difficult to digest. This leads to measurement variability, and poorer precision for this sample type. This is further illustrated in the next section. Conversely, the precision of Petra MAX is less affected by this sample type due to the minimal sample preparation needed for this analyzer.
Figure 11: Vanadium Repeatability
Figures 15 and 16 are box and whisker plots of the study data for nickel and vanadium by HDXRF and ICP, broken out by sample. ICP sample distribution is shown in orange, and HDXRF sample distribution is shown in blue. Additionally, the gravimetrically doped concentration of sample A is plotted as a red dot to illustrate the expected value. The medians for both Petra MAX and ICP are very close to the gravimetric value, and are similar for crude oil samples B, C, and D, indicating good correlation between techniques. However, the variability of individual ICP results is clearly much higher than Petra MAX, indicating that Petra MAX has better precision than ICP. The box plots also demonstrate how bituminous crude oil (sample D) has a negative impact on precision, especially for ICP.
Figure 15: Nickel Box Plot
Figure 16: Vanadium Box Plot
Conclusion
Author Contact Details Kyle Kuwitzky, Senior Product Manager and Leslie Johnson, Applications Scientist • XOS • 15 Tech Valley Drive, East Greenbush NY, 12061, USA • Tel: 1.518-880-1500 • Email:
info@xos.com • Web:
www.xos.com
Kyle Kuwitzky Leslie Johnson
Sulfur has long been a critical quality parameter for crude oil. Changes in crude oil production technologies have resulted in an increased importance of monitoring nickel and vanadium. This study demonstrates that HDXRF shows good correlation with ICP for nickel and vanadium in crude oil. Petra MAX, powered by HDXRF, shows better precision and lower variability when compared to ICP for the crude oils studied. With minimal sample preparation and rapid results, Petra MAX is an ideal solution for petroleum laboratories.
OCTOBER / NOVEMBER •
WWW.PETRO-ONLINE.COM
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52
