Advanced Sulfur Analysis in Hydrocarbons

better analysis counts

ADVANCED ANALYSIS WITH HDXRF Petra is powered by High Definition X-Ray Fluorescence (HDXRF) technology: an elemental analysis technique offering significantly enhanced detection performance over traditional EDXRF technology. This technique applies state-of-the-art monochromating and focusing optics, enabling dramatically higher signal-to-background ratio compared to traditional polychromatic X-Ray Fluorescence. Figure 1 shows the basic configuration of HDXRF and its use of focused monochromatic excitation. In this system, the diffraction-based doubly curved crystal optics capture a wide angle of X-rays from the source and focus a narrow energy band (monochromatic) of X-rays to a small spot on a measurement cell. The monochromatic beam excites the sample and secondary characteristic fluorescence X-rays are emitted. A detector processes those secondary X-rays and the instrument reports elemental composition of the sample. Figure 2 compares the detector signal of polychromatic (competitor) with monochromatic (XOS) XRF to demonstrate how monochromatic excitation reduces background noise and improves signal definition, delivering lower limits of detection and dramatically better precision. HDXRF is a direct measurement technique that requires no sample conversion, equating to no consumable gasses, little to no sample preparation, and results in just minutes.

Figure 1: HDXRF Technology Figure 2: Superior Signal-to-Noise Ratio

PRODUCT SPOTLIGHT Petra MAX is a robust benchtop analyzer that complies with ASTM D4294 and ISO 8754 for measuring sulfur in hydrocarbons. Petra MAX is powered by HDXRF, utilizing XOS patented doubly curved crystal optics coupled with a high-performance silicon drift detector and an intense monochromatic excitation beam. This industry-leading technology reduces background noise and increases signal-to-noise output, enabling low detection limits and high precision without the need for consumable helium gas, a vacuum pump, or extensive sample preparation.

INTRODUCTION Petra MAX™ delivers advanced D4294 sulfur analysis in addition to 12 elements from P to Zn including Ca, Fe, K, Ni, and V at sub-ppm levels for various applications: Sulfur in diesel and gasoline using ASTM D4294 and

ISO 8754 Sulfur and heavy metals in crude using ASTM D4294

and ISO 8754 Sulfur, Calcium, Phosphorus, and Zinc in unused

lubricants using ASTM D6481 Sulfur in bunker fuel and marine diesel using ISO 8217Throughout this whitepaper, we will discuss the following application studies: Application Study #1: Avoid Sulfur Measurement Bias

from Particulate Settling in Crude Oil Application Study #2: S, Ni, V, and Fe Analysis of Crude

Oil Using HDXRF®

Application Study #3: Total Sulfur in Hydrocarbons by ASTM D4294

APPLICATION STUDY #1: AVOID SULFUR MEASUREMENT BIAS FROM PARTICULATE SETTLING IN CRUDE OIL

BACKGROUND Test methods for measuring sulfur content, like ASTM D4294 and ISO 8754, have become critical for assessing the value of crude oil. These methods utilize X-ray Fluorescence (XRF) analysis, and with any method, it is important to consider the interferences inherent within the analysis technique. The ASTM D4294 standard test method references matrix effects as a known interference, which may influence the sulfur measurement in crude oil and produce biased results.

CHALLENGE Crude oil presents unique challenges for ASTM D4294 analysis. While sulfur containing compounds in crude oil are primarily comprised of organosulfur compounds that remain homogeneous in hydrocarbons, interfering compounds containing elements like Si, Ca, Cl, and Fe are commonly present as particulate solids that settle to the bottom of a sample over time. These particulate solids can absorb the X-ray signal and reduce the concentration of sulfur reported. While many D4294 instruments (traditional XRF) can correct for interfering elements, particulate settling in crude oil can create challenging scenarios. Diagram A demonstrates particulate settling over a period of 60 minutes. Particulate settling in crude oil has shown to cause underreported sulfur measurements by as much as 40%. Such a significant error can cause misclassification of sour crude oil as sweet crude oil. With global regulatory trends lowering sulfur levels in refined products from diesel to marine fuel, underreporting sulfur may cause refiners to miscalculate the costs associated with processing incoming crude oil. Because D4294 instruments (traditional XRF) take their measurement from the bottom of the sample, particulate settling occurs at the focal point of the analysis rendering the analyzer’s automatic interference correction, ineffective.

Diagram A: Particulate Settling

2

SOLUTION Many D4294 analyzers are designed with the X-ray detector focused on the bottom of a sample cup where particulate settling occurs, as depicted in Diagram 1. Since particulates settle over time, it is difficult to obtain accurate sulfur measurements due to the changing concentration of interferences. To combat the effects of particulate settling in crude oil, Petra MAX delivers a new, innovative sample chamber that rotates the sample on its side, providing a clear measurement window for more accurate results. See Diagram 2.

PARTICULATE SETTLING STUDIES To evaluate the effects of particulate setting, crude oil samples were obtained from three North American refineries. The samples were received in five-gallon drums and then stored in one-liter containers. The iron concentration for each sample was used to estimate the degree of particulate settling. Table 1 shows a summary of the iron concentration and particulate classification for each sample.

Particulates

X-rayDetector

Particulates

X-rayDetector

Diagram 1: Traditional XRF

SAMPLE INTRODUCTION METHODS

Diagram 2: Petra MAX™

Table 1: Crude Oil Classification

Iron (ppm) Particulate Settling

Crude A 35 High

Crude B 8 Medium

Crude C 2 Low

Note: Organosulfur compounds are homogeneous in the sample. Particulates represent elements like Ca, Cl, and Fe.

TRADITIONAL XRF VS. PETRA MAX To study the effects of particulate settling on sulfur measurements, a crude oil sample was analyzed using a traditional XRF analyzer and Petra MAX. Refer to Diagrams 1 and 2 for sample introduction methods. The following sample analysis procedure was performed using both methods: A particulate-free certified reference

standard of 2 wt% S in mineral oil sample was measured for 100-seconds to check instrument accuracy

One-liter bottles of crude oil were shaken vigorously, and samples were prepared and measured immediately for 100-seconds

Measurements were repeated 5 times with a 5-second pause in-between. The data was collected and compiled to evaluate the effect of particulate settling on sulfur analysis.

APPLICATION STUDY #1: (CONTINUED)

3

RESULTS 2% S in Mineral Oil – No ParticulatesIn order to test the accuracy of each sample introduction method, a particulate-free certified reference standard of 2% sulfur in mineral oil sample was analyzed using both traditional XRF and Petra MAX. Results for both methods demonstrate excellent accuracy. No particulates were present, and all measurements were within 1% of the known sulfur value and met the repeatability requirements for ASTM D4294 and ISO 8754. These results show that in the absence of particulate settling, both sample introduction methods provide accurate results.

Crude B - Medium Level of ParticulatesThe results for Crude B, containing a medium (common) level of particulates, are shown in Graph 2. In this crude oil sample, the drift in sulfur concentration for traditional XRF analysis is much less than in Crude A. However, there is a 12% lower sulfur concentration reported by the traditional XRF analysis than Petra MAX, demonstrating that even medium levels of particulate settling still impact the reported sulfur concentration. Petra MAX delivers stable results over the five repeat measurements of Crude B.

Graph 1: Crude A Results

Graph 3: Crude C Results

Graph 2: Crude B Results

Crude C - Low Level of ParticulatesThe results for Crude C, containing a low level of particulates, are shown in Graph 3. These results demonstrate that when particulate settling is low, both the traditional XRF and Petra MAX methods show agreement in reported sulfur concentration. This confirms that particulate settling is the cause for underreported sulfur concentrations with traditional XRF analysis.

Crude A - High Level of ParticulatesResults for Crude A, containing a high level of particulates, are shown in Graph 1. While the traditional XRF results show a rapid drift in sulfur concentration due to particulate settling, the results from Petra MAX remain stable for each repeat measurement. This demonstrates that, even with high levels of particulates, Petra MAX delivers accurate and precise sulfur measurements in crude oil for ASTM D4294 and ISO 8754 methodology.

APPLICATION STUDY #1: (CONTINUED)

4

CONCLUSION In conclusion, crude oil samples with medium to high levels of particulate settling may cause a matrix effect interference of the sulfur signal with traditional XRF when using ASTM D4294 and ISO 8754 methodology. Because settling can happen very quickly, even rapid sample preparation and measurement cannot prevent underreported sulfur in crude oils that exhibit particulate settling. Increased availability of crude oils with properties at the extreme end of the API scale, like light shale oils and heavy crude from oil sands, has increased the blending of crude oils in order to attain desirable properties that match refinery-operating requirements. This study demonstrates that matrix effects from particulate settling can effect reported sulfur results by as much as 40% in traditional D4294 analysis, which will likely lead to misclassifying sweet and sour crude oil.The new sample introduction technique utilized by Petra MAX eliminates the matrix effects altogether. As demonstrated throughout this study, Petra MAX delivers stable results regardless of the level of particulates in the crude oil.

RESULTS SUMMARY Table 2 below shows a summary of the total sulfur drift results of all three crude oil samples from the first to the fifth 100-second measurement, after sitting for 500-seconds. Results from the particulate-free reference standard samples are also included.

When comparing the results for the particulate-free certified reference sample (2% S in Mineral Oil) between Petra MAX and traditional XRF, there is no drift or bias present. When comparing results for the Crude A sample, there is a significant difference in the reported sulfur concentration. In the initial measurement (repeat #1) for Crude A, the traditional XRF analysis reported 26% less sulfur than Petra MAX. This demonstrates that even if samples were prepared and measured quickly, traditional methods still significantly underreport the sulfur concentration. After the fifth measurement (repeat #5) for Crude A, the traditional XRF analysis reported 42% less sulfur than Petra MAX.

Table 2: Total Sulfur Drift Results – Petra MAX vs. Traditional XRF

CRUDE A High Level of Particulates

CRUDE B Medium Level of Particulates

CRUDE C Low Level of Particulates

2% S in Mineral Oil No Particulates

Repeats (100s) Petra MAX Traditional

XRF Petra MAX Traditional XRF Petra MAX Traditional

XRF Petra MAX Traditional XRF

#1 0.930 0.690 0.850 0.751 1.304 1.301 2.010 1.995

#5 0.925 0.541 0.848 0.734 1.305 1.285 2.012 1.989

% Drift 0.5% 21.6% 0.2% 2.3% -0.1% 1.2% -0.1% 0.3%

All values for Sulfur in wt%

APPLICATION STUDY #1: (CONTINUED)

5

APPLICATION STUDY #2: S, NI, V, AND FE ANALYSIS OF CRUDE OIL USING HDXRF®

BACKGROUND Test methods for measuring sulfur content, like ASTM D4294 and ISO 8754, have become critical for assessing the value of crude oil. The blending of crude oils from different sources has become more commonplace within the industry to meet specifications for the classification of sweet crude oil. The introduction of new crudes brings new challenges, like higher concentrations of metals such as nickel (Ni), vanadium (V), and iron (Fe).

Ni and V are known to rapidly deactivate process catalysts in the fluid catalytic cracker (FCC) and hydrotreaters. This occurs because nickel and vanadium are often contained in large porphyrin molecules, which are not able to penetrate into catalyst pores. As a result, the nickel and vanadium containing molecules end up depositing on the catalyst and plugging the pores, which blocks the active sites located within the catalyst material. There are many pretreat options for addressing Ni and V, however these systems require accurate understanding of the Ni and V content to appropriately treat these metals. In response, many refiners have incorporated Ni and V analysis into their routine crude assay, and pipelines have set specifications for Ni and V in their common stream sweet crude.

Fe is introduced into crude oil from corrosion byproducts during transportation and can lead to pump and exchanger fouling, and off-specification coke.

At most refinery and test labs, analysis of S, Ni, V, and Fe are performed separately. The sulfur analysis is performed using EDXRF and metal content is identified using ICP-OES. While EDXRF is capable of measuring Ni, V, and Fe in addition to S, the limits of detection do not meet the levels needed to control refinery processes. ICP-OES is able to provide the needed sub-ppm analysis, however sample preparation is complex and takes many hours to complete. It involves a series of heating, acid digestion, and ashing procedures that are quite labor intensive.

A rapid measurement technique for sulfur compliance and simultaneous analysis of Ni, V, and Fe is necessary to meet the needs of refiners, pipelines, terminals, and other petroleum test labs.

APPLICATION STUDY In each study, ten separate aliquots were prepared and analyzed for 5 minutes each. Their individual measurement results and average are reported. Samples were prepared by transferring 6 mL to a 43 mm XRF sample cup and sealed with an Etnom film. Each refinery or pipeline location has their own specifications for specific elements like S, V, Ni and Fe. The desired level or limit for each heavy metal may vary depending on the detriment its presence causes to the equipment, process, or finished product. But in the case of V, Ni, and Fe, current methods can take a significant amount of time to prepare, including hours for ashing and analysis, and if outsourced, can be quite costly. Table A outlines common pipeline feed specifications.

ACCURACY STUDY To study the accuracy of Petra MAX, ten repeat measurements were performed on a commercially-available mineral oil reference material containing 10,000 ppm of S, and 10 ppm of V, Fe, and Ni. See Table 1 for results.

Table 1: S, V, Fe, & Ni in Mineral Oil (ppm)

Expected (ppm) 10,000 10 10 10

Repeats S V Fe Ni

1 9,867 11.1 9.3 10.2

2 9,854 11.3 9.2 10.1

3 9,707 11.2 9.0 10.0

4 9,662 10.8 9.1 10.0

5 9,739 10.9 9.0 10.1

6 9,727 11.1 9.0 10.1

7 9,656 11.1 9.0 10.1

8 9,634 10.9 9.0 10.1

9 9,636 10.8 9.0 10.0

10 9,671 11.0 8.9 10.2

Average 9,715 11.0 9.1 10.1

Standard Deviation 84.5 0.17 0.12 0.07

RSD% 0.9% 1.5% 1.3% 0.7%

Table A: Common Pipeline Feed Specifications

Element Specification (ppm)V < 5Ni < 5

V & Ni < 5Fe < 7

6

Table 2: S, V, Fe & Ni in Crude Oil Sample A (ppm)

Repeats S V Fe Ni1 1,057 1.03 0.37 0.092 1,082 1.04 0.42 0.163 1,071 1.10 0.35 0.174 1,067 1.10 0.41 0.085 1,059 1.10 0.38 0.156 1,062 1.04 0.37 0.137 1,081 1.07 0.33 0.178 1,083 1.08 0.42 0.119 1,085 0.96 0.48 0.1010 1,047 1.06 0.33 0.16

Average 1,069 1.06 0.39 0.13Standard Deviation 13.1 0.04 0.05 0.03

RSD% 1.2% 3.8% 12.8% 23.1%

Table 3: S, V, Fe & Ni in Crude Oil Sample B (ppm)

Repeats S V Fe Ni1 4,716 0.35 0.51 2.502 4,752 0.35 0.42 2.473 4,756 0.31 0.56 2.554 4,833 0.41 0.57 2.575 4,750 0.36 0.51 2.516 4,690 0.32 0.47 2.517 4,786 0.30 0.50 2.578 4,721 0.32 0.49 2.559 4,793 0.27 0.51 2.5610 4,749 0.31 0.49 2.52

Average 4,755 0.33 0.50 2.53Standard Deviation 41.4 0.04 0.04 0.03

RSD% 0.87% 12.1% 8% 1.2%

Petra MAX LOD ppm (600 second analysis time)

S V Fe Ni

5.70 0.10 0.07 0.04

APPLICATION STUDY #2: (CONTINUED)

PRECISION STUDY To study the precision of Petra MAX, ten repeat measurements were performed on two different crude oil samples containing S, V, Fe and Ni. The results shown in Tables 2 and 3 demonstrate that Petra MAX delivers precise measurement results well below desired specifications, and therefore is a valuable tool for monitoring trends as well as identifying materials that simply do not meet specification.

LIMIT OF DETECTION By incorporating patented doubly curved crystal optics to monochromate and focus the excitation beam, Petra MAX is able to achieve low limits of detection without the assistance of a vacuum pump or consumable helium.

CONCLUSION In response to the increased blending of sweet crude oil with crude containing higher levels of metals and S, petroleum labs are beginning to see the need to measure other elements, in addition to S, as a part of their crude assay. This study demonstrates that Petra MAX delivers simultaneous trace metals and ASTM D4294 or ISO 8754 analysis of S, in a single measurement. Petra MAX performs this analysis without the difficult ashing sample preparation that makes the analysis of metals like V, Ni, and Fe in crude oil so difficult today.

7

1.518.880.1500 • info@xos.com • xos.com

APPLICATION STUDY #3: TOTAL SULFUR IN HYDROCARBONS BY ASTM D4294

BACKGROUND Recent international trends across a majority of fuel types have focused on lowering specifications for total sulfur. These changes in regulation are in response to public health and environmental concerns. Examples include:

The International Maritime Organization’s announcement that MARPOL Annex VI, a regulation intended to control airborne emissions from ships, will reduce the total limit on sulfur content of fuel oil used on ships from 3.5 wt% to 0.5% by January 1, 2020.

In India, all automotive diesel and gasoline was transitioned to Bharat IV with a 50 ppm maximum sulfur in April 2017.

Jet fuel specifications such as ASTM D1655 and D6615 require a maximum sulfur of 3000 ppm.

To demonstrate compliance with these regulations, refinery and independent laboratories have long relied on standard methods such as ASTM D4294 and ISO 8754. These labs operate in fast-paced environments that require instruments that can provide rapid, precise results and are increasingly relying on solutions that provide simple, direct integration with laboratory information management systems (LIMS).

APPLICATION STUDY In this study, various hydrocarbon matrix samples were analyzed to demonstrate the expected performance of Petra MAX utilizing ASTM D4294 methodology. Separate aliquots were prepared and analyzed for 5 minutes each. Their individual measurement results and average are reported in the tables below. Samples were prepared by transferring 6 mL to a 43 mm XRF sample cup and sealed with an Etnom film.

ACCURACY STUDY To study the accuracy of Petra MAX, ten measurements were performed on a commercially-available mineral oil reference material containing 20,000 ppm of S. Samples were run on four analyzers by five different users. Results can be found in Table 4, and demonstrate the accuracy achievable with Petra MAX.

Table 4: S in Mineral Oil (ppm)

Expected (ppm) 20,000

Analyzer 1User 1 20,559

User 2 19,301

Analyzer 2 User 3 20,522

Analyzer 3 User 4 20,126

Analyzer 4 User 5 19,664

Average 20,034

Standard Deviation 547

RSD% 2.7%

PRECISION STUDY To study the precision of Petra MAX, various hydrocarbon samples were analyzed. These samples include heating oil, kerosene, jet A, vacuum gas oil (VGO), and crude oil. The analyses were performed at separate laboratories with unique instruments and users. The results shown in Table 5 demonstrate that Petra MAX delivers precise measurements across a wide range of hydrocarbon sample types. In addition, Petra MAX is a valuable tool to monitor trends and identify materials that simply do not meet specification.

CONCLUSION Sulfur analysis using ASTM D4294 and ISO 8754 continues to be an important measurement for refinery and independent laboratories around the world. Global regulation trends toward lower sulfur fuels demonstrate the need for a rapid and precise analysis solution. That solution is Petra MAX. Utilizing the most advanced optics technology, Petra MAX delivers accurate measurement across various hydrocarbons without the need for complex sample preparation or consumable gasses.

XOS. All rights reserved. HDXRF, Petra MAX and Petra 4294 are trademarks of XOS.

Table 5: S in Hydrocarbons (ppm)

Sample Type Crude Oil VGO Kerosene Heating Oil Jet A

Analyzer 1 User 1

16,553 6,754 146 391 93016,509 6,851 148 405 906

Analyzer 2 User 2

16,353 7,003 148 410 97616,751 7,017 152 428 974

Analyzer 3 User 3

17,666 7,002 161 436 98717,415 7,202 157 430 1,023

Analyzer 4 User 4

17,219 7,284 157 435 1,02817,382 7,308 138 434 1,017

Average 16,981 7,053 151 421 980Standard Deviation 497 199 7 17 44

RSD% 2.9% 2.8% 4.9% 4.0% 4.5%