When there is a need to monitor a process variable that extends beyond the “big four” (pressure, flow, temperature, and level), some type of process analyzer usually does the job. And this category of instrumentation has its own big four subtypes: Since manufacturing processes are generally designed for specific functions, the composition or characteristics of a given product should...
When there is a need to monitor a process variable that extends beyond the “big four” (pressure, flow, temperature, and level), some type of process analyzer usually does the job. And this category of instrumentation has its own big four subtypes:
Composition—Detect and measure specific chemical components in the process stream;
Electrochemical—Measure specific ion concentration, most commonly hydrogen (pH);
Spectrophotometric—Use light absorbing characteristics to detect and measure specific components; and,
Physical property—Measure specific gravity, density, viscosity, etc.
Since manufacturing processes are generally designed for specific functions, the composition or characteristics of a given product should fall into relatively narrow bands. For example, while a broad spectrum analyzer that can break down any unknown substance into its component parts may exist in a lab (at least on TV crime dramas), such are not generally practical in a real-life production environment. Gas chromatograph and mass spectrometer devices can quantify a wide range of substances, but their cost and relative complexity make them a choice when simpler technologies can’t do the job.
“Customers ask us about reliability more than accuracy or precision,” says Gary Brewer, product manager for ABB’s process automation division. “Manufacturers don’t usually have large support staffs, so if the analyzer goes down, the whole process can go down. So the rule of thumb is to apply the simplest technology possible that will get the needed measurement.”
When trying to quantify a component or contaminant in a product, it is critical to know what substance you’re looking for. In situations where more than one test has to be performed, more than one analyzer or analyzer technology will likely be involved.
Consider an example where the task is to analyze effluents in flue gas from a boiler, determining quantities of sulfur dioxide (SO 2 ), nitrous oxides (NOx), acid gasses, and mercury. SO 2 can be measured by infrared absorption, non-dispersive infrared, or ultraviolet. For NOx, use chemiluminescence or ultraviolet. Mercury calls for ultraviolet. Acid gasses may require a gas chromatograph or mass spectrometer. Even though ultraviolet may work for three of the four, the situation will likely require more than one sensor, perhaps even a separate sensor for each substance. Multi-component sensors are available, but there are trade-offs with cost and complexity.
Some relatively simple tests, such as pH or dissolved oxygen, can be handled by a probe inserted into a process stream. However, few analysis technologies are this simple. Most involve moving a sample of the product to a device where it can carry out a more complex analytical operation. Analysis becomes, in effect, a batch operation where a sample is extracted from the process stream and checked, just as if it were carried to a lab manually. This process can be automated and happen at appropriate intervals. Methods for designing piping to move samples to the analyzer is another issue. (See article in July Control Engineering .) In some cases to save cost, one device can serve more than one process line, as long as it can handle the span of variables required. Technologies have their own cycle time for analysis. Near infrared is fast enough that it can be essentially continuous, while a gas chromatograph can take 30 minutes to complete a test.
The key to selecting an analyzer, like any piece of instrumentation, is to understand your process and what information is critical to the larger control strategy.
Peter Welander is process industries editor. Reach him at PWelander@cfemedia.com .