Foremost among the areas in a refinery that are critically monitored for regulatory and consent decree compliance are the complex and challenging environments associated with the primary production processes, storage and waste disposal areas, including tank farms, additive blending, sour-gas stripping areas, fluid catalytic cracking units , secondary units to thermal oxidizers, vapor sumps and sewers. Such areas are characterized by complex mixtures of VOCs often at high concentrations because of their proximity to source areas, and by serious unpredictability due to the dynamic nature of the involved processes.

Monitoring such areas is difficult. Often the first indication of a problem is an alarm from a perimeter monitoring or fence-line system. Such systems, whether based on optical or other sensor technologies, coupled with atmospheric monitors can quickly direct crews to likely problem areas. Additionally, regular maintenance monitoring of control or containment systems at source locations is conducted to gauge regulatory compliance and associated trending. In such areas, the principal concerns center on the TVOC levels and benzene emissions, since both are the primary focus of regulatory organizations and consent decrees.  

Monitoring with PIDs

The generally accepted instrument used to monitor benzene levels in a refinery is a PID with a pre-filter tube. These tubes react with or block VOCs other than the compound of interest, e.g., benzene. This technique is the inverse of the long-established, direct reading tube technology; the color change or gradient is now used to indicate tube life rather than a quantitative value. Outflow from a pre-filter tube is measured by a PID with the assumption that readings will only occur if the compound of interest is present.

Pre-filter tube manufacturers utilize a variety of chemical reactions depending on what particular compound is being monitored. Reduction-oxidation reactions are commonly used to support benzene detection as well as identification of many other organics. Oxidizing agents (i.e., protonic acids) such as a chromate (Cr6+) are used to form a reaction barrier for compounds such as unsaturated aliphatics or substituted aromatics, allowing benzene to pass through for detection purposes.

All pre-filter tube manufacturers caution users to be very aware of the presence of possible interferences in the test environment to ensure accurate readings. Interference compounds can cause two significant levels of inaccuracy relative to the true presence of benzene. High readings or over-reporting can be caused by similar class compounds that are present in the sample and are not captured by the tube’s reactants; or by exceptional TVOC levels that overwhelm the tube’s limited contents and break through to the detector. Low readings or under-reporting can be caused by interference compounds that create a chemical or physical barrier in the tube to the compound of interest, thus reducing detection; also, unreacted alkanes can pass through the tube and inhibit detector response.

Over-reporting may result in pre-mature maintenance work that is costly and can unnecessarily disrupt on-going processes. Under-reporting can expose refinery operations to the punitive and costly consequences of non-compliance with regulatory or consent decree requirements.  

The Texas test

Enlarge this picture 1. Petro Pro GC from Photovac

To better understand the nature of the complex environments associated with refining processes, Photovac staff surveyed over 50 test sites at several refineries in southeastern Texas. The samples were analyzed with a portable gas chromatograph (GC) and also tested by an independent testing laboratory to determine the TVOC levels and the range of VOCs present. The laboratory analyses were performed according to EPA Method 18 (methane and ethane in air by GC/FID) and EPA Method TO-15 (VOCs in air by GC/MS). Several sites were repeatedly sampled over several weeks to better understand the dynamic nature of the associated processes.

Highlights of the test results included:

• TVOC levels ranging from 45 ppm to 240,000 ppm
• Several classes of interference VOCs including:
  • unbranched alkanes;
  • branched alkanes;
  • aromatics;
  • alkenes (straight and branched chains);
  • cycloalkanes;
  • alcohols and ethers;
  • and sulfur compounds (straight or branched chain)
• Alkanes (butanes & pentanes) often in excess of 60,000 ppm
• Cycloalkanes (cyclohexane) often in excess of 1100 ppm
• Alcohols and ethers (ethanol, TAME) in excess of 250 ppm

When trying to monitor such test sites there is no way to completely know in advance all the contributing factors to the sampling environment – particularly sewers, sumps and waste water wells. As such, one cannot have an accurate up-front understanding of the presence of potential interferences or cross-sensitive compounds. And where the composition of the effluent stream might be known, total concentrations cannot be accurately anticipated.

It is no surprise then that the PID with a pre-filter tube, usually accurate in the dilute perimeter or fence-line areas, can be erratic in more intense environments. GC measurements of the test sites at the Texas refineries, when compared to the PID plus pre-filter tube, clearly demonstrated the distorted readings that can occur because of the limitations of testing based on chemical reactants. A comparison of the two technologies is shown in Figure 1.  

Understanding the technology

To better understand how a GC works, it is helpful to know that chromatography is analogous to industrial distillation in a vertical column with a series of perforated plates inside. In a distillation column, the plates are positioned as separation areas for various fractions (liquid, vapor or gas) to enter into equilibrium. The longer the column, and with more added plates, the better the separations. The height of the plates resolves to the distance along a gas chromatography column that gives the same separation as plates in a distillation column. The process collects the separation of chemicals at the plates, and it flows with the carrier gas, allowing complex mixtures to be separated in a gas chromatography column.

In the GC column, an air sample containing a mixture of VOCs is moved by a continuous stream of carrier gas (usually ultra-zero air or another inert gas) from the sample inlet through the column where separations occur. The components in the mixture separate as they move through the column due to differences in their rates of interaction with the column material. GC columns have the ability to handle a variety of compounds and potential interferences including the presence of compounds with characteristics similar to the compound of interest as well as high concentrations of compounds that might obscure detection of the compound of interest.

The separated compounds are measured by a detector as they elute or exit from the column. The length of time that any particular compound takes to pass through the column, called “retention time,” is unique for each compound.

A plot of the detector response versus time results in a chromatogram chart showing an individual compound’s identity (by retention time) and concentration (by peak area).  

Making it portable

Enlarge this picture The above graph compares the results read from three testers at a test site.

The problem with GC work, at least until recently, is that it has always needed a laboratory environment. While very accurate, lab work requires highly trained staff and can be expensive – especially if lab schedules and/or priorities are disrupted. Also, test results can be vulnerable to handling errors in transit between the test site and lab, and the process does not support timely response to problems by field maintenance staff.

Recently, companies have developed portable, or on-site monitors for certain gases, including, notably, benzene. This has made it possible for end users to employ portable GCs (such as the PetroPRO GC) for monitoring benzene in the primary production processes of refineries.