After some good debate in a LinkedIn forum, I came to the conclusion that this practice should not be called an “industry practice” as there were a lot of people in the industry that did not do it and had never heard of it.
It certainly is not part of any Recognized and Generally Accepted Good Engineering Practice (RAGAGEP) as not one person could provide such language in a RAGAGEP from the NH3 refrigeration industry (e.g. IIAR, ASHRAE). Someone did contact some manufacturers and all those he contacted said their valve(s) could be used either way, but when I asked them why they put the flow arrow on the valve housing the crickets began to sing. When asked if they would provide me a formal letter from their engineering group stating that I could install their valves backward and not void my warranty and they would stand with me if the valve leaked and I was cited by OSHA or EPA. I am still waiting for the return phone call, of course, I think they know what is going on now and why I was asking.
So when asking all my engineering contacts about this practice and if it is acceptable in their plants, it was unanimous of the 20+ that responded that I was crazy for even asking such a thing. Some thought it was a trick question and tried to figure the angle I was coming from and why would I even want to do such a thing. If I need to install a valve backward, then I’m most likely spec’ing out the wrong valve for the application.
So my wandering mind began to think…how would a plant manage such a program?
Would they show the valves that are installed backward on P&IDs?
What is the symbol to show a backward valve on a P&ID?
Does the facility indicate which valves are backward in their PSI?
Do they indicate it on Work Orders associated with the valves?
Is this trained on as part of operating and maintenance procedures?
How are contractors informed of which valves they install are to be installed backward?
Is part of an engineering spec as to which valves are installed backward?
What about PHA’s, are their failure modes with backward installed valves that are not typical causes of NO FLOW, TOO MUCH FLOW, etc.
Yesterday I went to one of my local clients and asked their four operators on 1st shift which of their valves are installed backward. Got blank stares. The second shift, same result with the two operators.
So I went to my OSHA and EPA sources and asked a simple question so as to not slant their view and give away the answer I think is the right answer:
“If during an inspection you came across a valve installed backward, would it be an indicator of a potential issue?”
Most of them asked how would they see this since most valves in an NH3 process are insulated, but if they came across this, yes it would be an issue that would need to be addressed.
What would the facility need to have on hand to wash away any concerns that this is was done per design and not by accident? Almost all of them went down the path that I tried to go down in the ammonia discussion forum. Those CSHOs who do PSM asked a lot of the same questions I had about how to identify these valves in procedures, P&IDs, PSI, etc.
Bottom Line…Make absolutely certain that the valve is being used as designed to be used by the manufacturer. If it is not perfectly clear that the valve can be installed backward, then the facility should ask for a formal letter from the manufacturer that allows their valve to be installed backward. The facility needs to have some means of identifying which valves are installed backward to ensure they are maintained in such a manner “as designed”. A few of them offered up some type of program like a car seal program identifies those valves; however, they all admitted they have never seen such a program in their careers.
So the debate goes on. Can valves be installed backward? If the manufacturers say yes and the facility meets all the manufacturer’s requirements AND the facility has some type of management program to ensure personnel knows which valves are to be installed backward, then yes it appears it may be safe to do.