February 15, 2011

Field-Packaged Gas Chromatographs

I am Carolyn Snyder, business development manager for the process industry at Rosemount Analytical.

For many years, all process gas chromatographs were installed in shelters that usually cost significantly more than the analyzers themselves. From a technician’s point of view, a shelter was desirable for its climate control and comfort to troubleshoot and repair the analyzer. However, using a fully enclosed shelter requires several compromises. First, you give up proximity to the point of measurement, which means you spend more on cabling and sample tubing (often heated). Second, a shelter is limited in the number of units it can house. As your needs change, you may be required to invest in additional expensive shelters to accommodate the changes in your analytical scope. Because of these issues, most major suppliers of GCs have produced some form of a field-packaged unit that is theoretically capable of existing in a non-climate-controlled environment.

Which is why I was quite surprised to hear many of these suppliers still strongly advocate the use of conventional analyzers at the recent IFPAC conference held in Baltimore, Maryland. Sitting through several papers, I listened to presenters explain that only the simplest measurements can be made with field-mounted units. They also discussed some service-related issues centering on the expense of the modular concept, where the entire module is replaced rather than repaired in the field, which can be very costly.

What I’ve discovered over the past few months is that none of these issues apply to the field-mounted gas chromatographs manufactured by Emerson (sold as Rosemount Analytical into the process industry and Daniel Danalyzer into the natural gas industry).

First, the Emerson GCs all go through an environmental chamber to ensure there is no failure when temperatures vary from 0° to 130°F. They are designed with access to the various parts – electronics, boards, columns, detectors, etc. – so that repair in the field is feasible. Most importantly, these field-packaged analyzers are not restricted to simple applications. Some of the most strenuous measurements are easily performed by these units which house multiple valves, columns and detectors in addition to stream switching capability. The measurement of C9+ with HHV (higher heating value, or BTU) and HCDP (hydrocarbon dewpoint) has been provided in scores of field analyzer units to the satisfaction of many major hydrocarbon processing companies.

In short, performance does not have to be sacrificed in a field housing analyzer if it is properly designed and implemented.

February 1, 2011

pH Detectives – What is Going On in the Process?

Hi, I’m Jim Gray. I kicked off this blog back in April. I read Shane Hale’s blog from this last July, “Sample Handling — Preparing for a Bad Day,” in which he emphasized the importance of taking process upsets into account when designing sample systems, and it got me thinking…

Far more often than not, pH sensors are simply inserted or submerged directly into the process without the benefit of sample handling. If the pH electrode and the liquid junction of the reference electrode can remain in contact with the process, and there are no components in the process which will dissolve the glass pH electrode, severely coat the sensor, or poison the reference electrode, all is good. This is the case in the vast majority of pH applications.

But, sometimes pH sensors are installed in processes with conditions and compositions that can shorten sensor life or result in too much required maintenance. This can happen during normal conditions or during upset conditions, but in either case, that’s when we get the call (or e-mail).

So what do we do? It is as simple as going back to square one and looking at the process conditions and composition, and yes, upset conditions. Temperature, pressure and flow data is usually easy to come by, but composition data seems a little harder to get. Sometimes the process contains a brand name product, like Acme SuperClean XYZ – getting the composition is a simple matter of looking at an MSDS sheet. 

Also of great help is diagnostic information from the plant historian and records — what alerts have been showing up? If the sensor will no longer calibrate, what calibration error alert is it showing? Additional information is helpful like: How long was the measurement good? Does failure occur during an upset or when the process is idle? Armed with this information we can usually find the root cause of the problem fairly quickly, and in most cases find a solution by recommending a sensor designed to deal most effectively with the issue, like high temperature, coating, reference poisoning, attack on the glass electrode, etc.

Here is an example where a little information goes a long way:

Caustic concentration was being measured using pH around 13.7 pH (2% NaOH).

The problem: The pH on-line doesn’t match the lab and the electrodes do not
last that long.

Just from a description of the application, we know the pH is highly temperature-dependent at this pH, and unless the sample in the lab is at the same temperature as it is in the process, they will not agree, even if everything else is OK, which it is not. There can also be junction potentials and sodium error involved, not to mention that NaOH can drastically shorten pH electrode life. Deriving NaOH concentration from pH involves the exponential of pH, and so any error in the pH will result in large errors in derived concentration.

The solution: Use toroidal conductivity for measuring strong acids and bases such as this; there will be far less maintenance, and a much more accurate measurement of concentration.

But all too often, the needed information is slow in coming, and as a result, so is the solution. Here is an example that I have to laugh at due to the blind alleys we went down until we got the whole story:

The problem: pH electrodes are breaking within a day or two in a caustic scrubber. A little later we are given the pH range which indicates that there is not enough caustic to harm the electrodes, and there are no upsets. A week or so later we find out that hydrogen fluoride is also being scrubbed, so we do the math and determine that at the pH and fluoride concentration, glass attack by hydrofluoric acid is not a problem. A week or so later we are told that some of the pH electrodes are already broken while still in the box — the problem has nothing to do with the process! We are then told that all of the electrodes were stored outside over the winter in northern Norway — they all froze.

Getting to the final conclusion took about 3 weeks. Had we had all the information up-front, we would have had the answer immediately.

Dealing with pH issues is sometimes like being a detective in a mystery novel — hence the title of this blog. What makes a mystery novel entertaining is the suspense of not knowing “who done it.” But, unlike a mystery novel, no one really needs the suspense when it comes to problem pH applications; it just prolongs maintenance costs and aggravation.

Getting all the facts in chapter one would make for a lousy mystery novel, but will get pH issues resolved a lot more quickly.