Get fast D4294 sulphur results without centrifuging

Petroleum professionals looking to certify their products are commonly tied to specific method types for analysis.

Joseph Iaia and Kyle Kuwitzky
XOS

Viewed : 1426

Article Summary

Two known methods, ASTM D4294 and ISO 8754, utilise a measurement technique known as X-ray Fluorescence, or XRF. Many methods come with their own unique set of interferences and bias corrections. For XRF, a common matrix effect interference involves particulates settling to the bottom of the sample cup and absorbing the X-ray signal. This matrix effect will ultimately influence the total sulphur measurement and lead to a biased result. Refineries and third-party certification companies using XRF for high-particulate samples have relied on centrifuging samples to ensure an accurate sulphur measurement – which involves a time-consuming sample preparation process.

Challenge
While many D4294 instruments (traditional XRF) can correct for interfering elements, interferences that settle in crude oil can create challenging scenarios. Diagram A demonstrates settling over a period of 60 minutes.

Particulate solids and water have shown to cause underreported sulphur 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 sulphur levels in refined products from diesel to marine fuel, underreporting sulphur 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, settling occurs at the focal point of the analysis rendering the analyser’s automatic interference correction, ineffective. To prevent biased results, many laboratories centrifuge all crude oil samples prior to analysis by traditional D4294 instruments. This increases the amount of processing and time it takes to perform the measurement.

Solution
Many D4294 analysers are designed with the X-ray detector focused on the bottom of a sample cup where settling occurs, as depicted in Diagram 1. Since particulate solids and water settle over time, it is difficult to obtain accurate sulphur measurements due to the changing concentration of interferences. To combat the effects of 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.

In the following whitepaper we will discuss how to eliminate the need to centrifuge for D4294 sulphur analysis in two different application studies:
• Application Study 1: Obtain Accurate Sulphur Results Without Centrifuging in Real-World Samples
• Application Study 2: Petra MAX vs Traditional XRF for D4294 Sulphur Analysis

Application Study 1:
Obtain accurate sulfur results without centrifuging in real-world samples
In the following application study, we will assess the ability of Petra MAX™ to eliminate the need to centrifuge by running a real-world crude oil sample supplied by a large pipeline. In this paper we will run these samples both with and without centrifuge pretreatment. We will also run the same samples using a traditional XRF analyser setup, with and without centrifuge pretreatment, and compare the results.

Experiment
The sample used for this experiment contained an iron concentration of 28 ppm. Samples were prepared by pipetting roughly 8mL of sample into a standard XRF cup after it was dusted off with canned air. Then the cup was sealed using a sheet of Etnom® film which was also dusted off with canned air. Prepped samples were then placed in their respective XRF analysers and measured for 100 seconds for 3 repeats. We ran each sample 3 times to demonstrate the effects of particulate settling over the duration of the three measurements.

Each sample type was centrifuged prior to preparation of a new sample at a relative centrifugal force (RCF) of 600 for 60 seconds. After, samples were measured for 100 seconds for 3 repeats with both analysers.

Note: No significant difference was seen in vented versus non-vented samples, as is common with this type of application.
Results
Understanding the data
The data gathered from the experiments was compiled into several different graphs that showcase various findings. Each graph will be shown and explained with additional data sets. First, we will explore the effects of iron settling in a sample by showing non-centrifuged results on a traditional XRF system. Then we will incorporate data from traditional XRF techniques after centrifuging. Lastly, we will bring in data run on a Petra MAX.

A complete, comprehensive graph (Graph 4) can be found at the end of this segment.

The effects of settling with traditional XRF
The data shown in Graph 1 indicates that there is a direct relationship between the iron concentration, marked by the dotted line, and sulphur concentration when using traditional XRF, marked by the solid orange line as the sample settles over time. The takeaways from this data are:
• As iron settles in the sample cup, sulphur concentration goes down.
• This bias may lead to underreported sulphur.
• Iron settles at different rates.
• With traditional XRF, it is difficult to account for this iron settling without specific sample preparation.

Mitigating bias by centrifuging with traditional XRF
We concluded from our first graph that accounting for this bias without conducting additional sample preparation is difficult in traditional XRF analysis. When using traditional XRF, this is done by centrifuging the sample, which adds preparation. In Graph 2 we have added in sulphur data for the same sample after centrifuging on a traditional XRF analyser. The following can be observed:
• The sulphur concentration has mostly stabilised at a higher concentration than the non-centrifuged sample.
• Removal of particulate matter by centrifuging (including iron) mitigates the bias issue.

Removing the need for centrifuging with Petra MAX
As we have observed, centrifuging is necessary when using a traditional XRF system. What might the data look like when using a Petra MAX analyser and its vertical sample introduction method? First, we will showcase Petra MAX data on a centrifuged sample in Graph 3, which acts as a study control to show that the centrifuged sample data is comparable whether run on Petra or a traditional XRF system.

Lastly, to showcase how Petra MAX can eliminate the need for centrifuging, we will display all the data together in Graph 4. This combined data brings in results from measurements of the same crude oil sample run on a Petra MAX without centrifuging. The Petra non-centrifuged data shows increased measurement stability over the traditional XRF data (both centrifuged and non-centrifuged) and a slightly higher bias over the centrifuged data (both traditional and Petra XRF).