Impurities Going Undetected?

How to ensure accurate measurement of carbon monoxide and carbon dioxide in polyethylene production

1 of 2 < 1 | 2 View on one page



Polyethylene manufacturers rely on ethylene producers to supply product that conforms to well-defined specifications. Three impurities are critical from a catalyst point of view: carbon dioxide (CO2), carbon monoxide (CO) and acetylene. The precise and accurate measurement of these three impurities is essential in determining the impact on the catalyst used to manufacture the polyethylene.

The catalyst is a major expense in polyethylene production; waste minimization can translate into substantial cost savings. However, current measurement technology might not detect trace levels of CO and CO2 accurately. CO can tie up eight times the amount of catalyst as ethylene alone. Nearly 50 billion pounds of ethylene is manufactured into polyethylene in North America each year, meaning this can amount to hundreds of millions of dollars in wasted catalyst.

With these high financial stakes, it is critical that the analytical method used to establish the background in ethylene is both accurate and precise. Therefore, the author's company* (the company) and the American Society for Testing and Materials' (ASTM) D2 Committee jointly sponsored a project that involved an extensive evaluation of the popular methanizer flame ionization detector (FID) gas chromatography (GC) method. The goals of this "round robin" project were:

To evaluate GC systems that analyze CO, CO2 and acetylene in an ethylene matrix.

To cover concentration ranges of 5 parts per million (ppm) down to 25 parts per billion (ppb).

To eliminate the impact of standards by providing two high-accuracy primary standards.

The results of this evaluation showed the popular methanizer FID technology had significant bias that could result in an underreporting of the level of CO and CO2; therefore, it could have a significant impact on the cost-efficiency of the polyethylene manufacturing process. These findings were presented at the Gulf Coast Conference last September in Galveston, Texas.

Experiment design

The overall experiment was designed in concurrence with representatives from both the company and the D2 committee. In general, each participant in the study received six different samples and two primary standards.

The company's advanced application group in Geismar, La., prepared both the samples and standards. The materials were prepared under a stringent protocol to ensure material accuracy.

The validity of the samples and standards was tested on two GC systems. Based on previous work, special attention was placed on the GC support gases, using only ultra-high-purity (UHP)-grade materials. Finally, the team applied a consistent and rigorous analytical protocol to material testing.

Analytical systems

Two analytical systems were used: a methanizer FID GC system and a pulse discharge ionization detector (PDID) GC system. The two systems were built around Agilent instrumentation modified by Wasson ECE. The data generated by these two systems were collected on an Agilent ChemStation system.

A few differences between the two systems were noted. The methanizer FID system was built on a 5890 Series II GC. The instrument used 5.0-grade (99.999 percent purity, or UHP-grade) helium (He) and hydrogen (H2) support gases. Separation of both CO and CO2 was accomplished on one analytical column.

The PDID GC system was built on a 6890 Series system. This system was equipped with a Wasson VPS system to enable sample pressure control, as well as a Valco Inc. PDID. The carrier gas was 6.0-grade (99.9999 percent purity, or UHP grade) He, followed by a high-temperature getter system. Separation of the CO and CO2 was accomplished on two columns.

Support gases

The quality of the support gases used for both carrier and detector makeup gases plays a significant role in measurement accuracy. For the methanizer FID tests, Test One used typical 5.0 grades of He and H2 as support gases. Zero-grade air was used for the FID support. Both He and H2 can have ppb levels of both CO and CO2 as impurities.

In Test Two, both He and H2 were changed to methanizer FID grade. In the case of H2, this grade is derived from a chloro-alkali process. Unlike that produced through other processes, this material is free from both CO and CO2.

In the case of He, the bulk material is purified through a cryosorption process. To further ensure the integrity of the gases, a high-temperature getter was placed in line for the He, and a low-temperature getter was placed in line for the H2.

In the case of the PDID system, a 6.0 grade He was used to prevent this universal detector from biasing both the CO and CO2 measurements. To ensure none of these compounds were present in the He, a high-temperature getter was placed between the cylinder and the GC system.

Preparing standards and mixtures

To ensure the validity of the tests, the researchers were required to control the following variables:

The quality of the ethylene used to make the mixtures.

The integrity of the cylinder in which the mixtures were to be prepared.

The methodology used to ensure the accuracy of the resulting mixtures.

The company acquired some very high-purity ethylene that was known to be low in both CO and CO2. This starting material was assayed to ensure the product claim. The cylinders used to make the test mixtures/standards were prepared to ensure the stability of both the CO and the CO2.

The sample and standard mixtures were prepared using the standard addition method and then analyzed. A more concentrated mixture was used to make the more diluted mixtures. To ensure the reliability of the resulting mixtures, a gravimetric uncertainty of +/-0.2 percent relative was used. Each mixture then was homogenized after the manufacturing process.

Each mixture was prepared based on the assumption that no CO or CO2 was present in the ethylene. The only CO and CO2 came from the blending mixture. The CO and CO2 background in the ethylene could be calculated based on a regression analysis of all the analytical results. The methane, ethane and acetylene concentrations were varied to represent typical systems.

1 of 2 < 1 | 2 View on one page
Show Comments
Hide Comments

Join the discussion

We welcome your thoughtful comments.
All comments will display your user name.

Want to participate in the discussion?

Register for free

Log in for complete access.


No one has commented on this page yet.

RSS feed for comments on this page | RSS feed for all comments