As I have discussed many times with many clients, the practice of “bench testing” relief valves (RV) is a common practice in many industries; but in some industries, the practice has been considered to be a waste of time and money. At least that was the case until now! It appears that one major trade group is taking a position that “bench testing” RV’s is not only a good safety practice, it is also GOOD BUSINESS. So maybe the days of hearing… “Bryan we are not a refinery… chemical plant… power plant… fertilizer plant… etc.” is behind us and we can move forward in improving process safety in the ammonia refrigeration industry. And yes I want to tip my hat and say THANKS to the International Institute of Ammonia Refrigeration (IIAR) for publishing “RVCalc, Using Weibull Analysis to predict valve replacement” this month (September 2014) in their “Condenser” magazine. We should not take this publishing lightly as it is a MAJOR step in the right direction and shows that IIAR is moving in the right direction. I can only imagine how this practice will be received by IIAR members, but without a doubt, it is a much-needed change in this industry. To provide some contrast to how many are currently managing their RVs, consider this…
It is time to change the RV’s. A worker grabs a trash bin and slides it over to the RV tree on a compressor. Following all his/her protocols they simply remove the RV from the tree and throw it into the trash bin. There is NO inspection to establish the “as found condition”, there may be some very loosely defined “examination” of the RV internals, but even that is often very subjective and done by an untrained/un-certified eye. The worker opens the packaging to the new RV and installs it per manufacturer’s instructions. It is more often than not, that simple and there lies the significant gap in process safety.
For far too long, facilities have read the “at least every five years” as “every five years”; obviously missing the key words “at least”. With these two tiny words included it becomes clear that five years is the max, but it also implies that there may be an occasion where five years is too long. So how would a facility establish their RV maintenance frequency? A practice called “Bench Testing” – where the RV is placed on a specialized testing device and tested to ensure it would function properly (i.e. as designed and set).
The ammonia refrigeration industry was all too quick to point out the business aspect of bench testing… a new RV cost $75. To send one out to get tested was $65. If it failed, they were out the $65 and still had to spend the $75 on the new RV. So let’s just cut out the testing costs, as we are going to install a new RV regardless of what the tests tells us. “After all, we are not a refinery or petrochem plant that has large and costly RVs that get rebuilt and recertified for reuse; it is just cheaper for us to install new ones”. Sounds like a good business plan, but a horrible process safety plan!
This month IIAR article makes an EXCELLENT business and safety case for why “bench testing” can be a good thing to the ammonia refrigeration industry. The article mentions a new software program that takes the bench test data and will analysis the data in order to establish an ENGINEERING BASIS for the frequency in which EACH RV on the process needs to receive any maintenance. This is just HUGE for this industry. Now if we can get Bulletin 110 revised so that the “practice” of replacing every RV(s) every five years, with NO inspection of used RVs and NO engineering basis for this schedule, is removed and not permitted we will have closed the gap entirely. We can only fight and work to make this happen.
It may take years to convince every refrigeration facility that “good engineering practices” call for an engineering basis to establish any frequency for each SPECIFIC process, but this is a step in the right direction and it is a big step with huge advantages and improvements to process safety within the ammonia refrigeration industry. I tip my hat to IIAR, Dr. Elder, and the University of WI for a job well done!
And while you are reading the RVCalc article, check out the update on the IIAR QRA project where they conducted a Quantitative Risk Analysis (QRA) on secondary safety systems associated with Relief Systems. Specifically, they looked at direct discharge to the atmosphere (simplest), diffusion tank(s), scrubber(s), flare (most complex), and emergency pressure control system(s). The article does not provide any data, but it is an INTERESTING project, to say the least, as the article states:
- the RV discharge to the atmosphere is the simplest and least likely to fail
- the diffusion tank has 8 failure modes
- scrubber has 15 failure modes
- flare has 18 failure modes