NUTS! Ammonia Refrigeration “RAGAGEP” Head Scratchers

In December 22, 1944, the allies (American troops) were sent a letter from the commander of German forces located outside of the town, demanding “the honorable surrender” of the town within two hours. General McAuliffe, the commander of the American troops reply was brief and to the point:

“To the German Commander:

N U T S!

The American Commander”

We, however, are obviously not in as severe a life or death situation (though it may seem that way), but occasionally we are asked or expected to do things that seem so utterly ridiculous that you can only say, this is NUTS! There are several of those when it comes to ammonia refrigeration systems. You have to wonder if those that developed the requirements have actually had to do this work in the field. This is not to be an indictment against those that developed the standards, but more so in the process at getting input from those that actually carry out the majority of the tasks. This goes to the heart of PSM “employee participation” and more importantly understanding why it is so vital. In fact, there is a good reason why it is the first of the 14 plus PSM elements.

Back in the day when I was a young engineer, I was fortunate enough to have a mentor that insisted that I spend at least 25% of my time working with plant operators to get to know what she called “the real process”, and not the one that was found on paper. The second bit of advice that I try to instill in those I mentor and coach today, is to design processes that are easy to maintain, test, and inspect. If you make things difficult, I can all but guarantee that they will NOT get done, shortcuts will be created that don’t achieve the desired results, cause the potential for injury, or they are simply “pencil whipped”. We need to spend time with those who are expected to do these tasks and requirements that we put in standards, and better yet, if possible, actually do them ourselves to understand what needs to be done and how. We need to do the things that help achieve process safety and not waste valuable resources by creating a requirement that, in most cases, will never be done as intended or more importantly, that achieves the desired process safety intent. So, what are those things that I believe are NUTS!

1. Nuts! The number of actual compressor shutdowns required to test all the cutouts: Depending on the set up, the normal number of shutdowns needed on average is six (6). Changing a set point is not a test unless the PLC logic has changed. Therefore, changing the set point to test a cutout only results in a partial test…though better than not tested at all. This one should get more than one “Nuts!” since most systems now are run through a PLC that can be configured for a "test mode" that could limit the actual shutdowns to one or two at most all while achieving a full functional test.

2. Nuts! Assuming that recognized and generally accepted good engineering practice (RAGAGEP) is that a pressure transducer can be calibrated without checking the zero. This is recognized and generally accepted bad engineering practice (RAGABEP).

3. Nuts! The inability to “easily” calibrate compressor pressure sensors. In some cases, in order to properly calibrate a pressure sensor (including zero pressure) a complete pump down of the compressor system is required. How often will this get done?

4. Nuts! Drilling holes in the insulation to conduct corrosion under insulation (CUI) or Non-Destructive Testing (NDT) inspections. This type of inspection is like trying to find a needle in the stack of needles.

5. Nuts! Not moving to or utilizing Risk Based Inspections (RBI) for vessels and piping. It is commonly known that he typical mode of failure of piping and vessels in ammonia refrigeration systems, is CUI, therefore, we should use resources on conducting higher risk inspection and understanding the mode of failure better and focusing resources on those areas where CUI risks are the highest.

6. Nuts! The number of inaccurate Relief System evaluations (header and inlet/out pressure drop) that result in both undersized and oversized relief devices and systems. Assuming the incorrect “P2” when using the “L” equation for just the relief valve piping to the header and not taking into account the enter header system to the outlet, or not taking into account back pressure on the system.

7. Nuts! Installing automatic valves on the outlet of a relief device(s) believing that just because it is not a “Manual” valve, that it is allowed by code.