RAGAGEP (Recognized and Generally Accepted Good Engineering Practices/Principles) is extremely important to our Process Safety programs as it helps define the boundaries of what is (and isn’t) acceptable in our processes and our management of them. There seems to be some confusion in a significant portion of the industry as to how to practically apply RAGAGEP* so I thought a brief discussion (and worked example) might be useful.
Let’s say that we have multiple possible RAGAGEP’s for a single item – such as relief valve replacement schedule. Those multiple RAGAGEP’s may well have differing requirements so we will need to rank them to understand what we actually need to do. Here’s an example RAGAGEP listing from OSHA:
- Codes adopted by the AHJ (Authority Having Jurisdiction) such as IMC/UMC
- Consensus Standards such as IIAR or ASHRAE
- Non-consensus documents such as Pamphlets / Bulletins from industry organizations
- Internal Standards such as your corporate policy
What isn’t on that list is manufacturer’s recommendations and there’s a reason why. The things listed above set the RAGAGEP and the manufacturer’s recommendations can modify it.**
There are generally two ways to modify something: make it more or less restrictive.
More: In the event that the manufacturer gives you a recommendation that is more restrictive (conservative) than the RAGAGEP, you must*** accept that more restrictive recommendation.
Less: If the manufacturer gives you a recommendation that is less restrictive than the RAGAGEP, you can accept that less restrictive recommendation, but you will need to document why you believe that the manufacturer’s recommendation is superior to the existing RAGAGEP.
In a recent article on RCE-Chill, we discussed the replacement schedule for a relief valve that relieves back into the system. The codes reference the consensus standards, which in turn reference some non-consensus bulletins. The bulletin in question, IIAR B110 says that these valves are not subject to the 5yr changeout frequency that other relief valves are. Yet, we have an email from Cyrus Shank’s engineering department that still recommends the 5yr changeout schedule.
In this case, we have a disagreement between the non-consensus bulletin and the manufacturer’s recommendation. Put another way, we have a generic recommendation on relief valve changeout versus a manufacturer specific recommendation. Obviously, the manufacturer’s specific recommendation on their valves overrides the generic recommendation about all relief valves. Therefore, as long as we are going to use these specific valves, we need to follow the manufacturer’s recommendation. ***
*It’s important to understand that we’re talking about what RAGAGEP decision is the most defensible during an inspection / audit.
** In 1910.119(j)(4)(iii) manufacturer’s recommendations are explicitly called out in conjunction with good engineering practices to set inspection/testing frequency, but the point still holds true.
*** It’s theoretically possible that you can make the engineering case that you know more about the manufacturer’s equipment as it operates in your process than they do, so you can override their recommendation. One method that’s routinely used is to choose an alternative way to achieve the same goals – one where you can show the engineering rationale to prove your alternative is as safe or safer. A common example of that would be replacing the oil based on regular oil analysis rather than changing it out at a specific hour interval. Of course, such a change would have to be thoroughly documented through your Management of Change procedure.