Monday, December 14, 2015

Accuracy should trump easy-peasy

In January the ASME A112.14.3 sub-committee will once again meet to try to update the standard, which has remained unchanged since its publication in 2000.

One of the issues that the committee has struggled to come to consensus on is the incremental test cycles and the skim tank procedure.

At the risk of boring the reader to death - something I normally try to avoid - let me explain the procedure and the inherent problem that I believe needs to be addressed.


This is what the ASME A112.14.3 test apparatus looks like, conceptually anyway.

Once the flow rate is calibrated from the test sinks to the desired flow rate for the interceptor being tested, the testing procedure involves opening the drain valve on the test sinks and "let 'er rip".

The flow from the test sinks will pass through the grease interceptor and into the skim tank, after which any accumulated lard found in the skim tank is skimmed off (hence the clever and sophisticated name of the tank), dewatered and weighed. The weight of the lard taken from the skim tank is deducted from the weight of the lard introduced into the test sinks and the difference representing the efficiency of the interceptor during that specific test cycle.

Now, the skim tank is supposed to have a waste outlet connected to the bottom of the tank at one end and be trapped to retain approximately 18 inches of water in the tank. The tank is also provided with a four (4) inch bottom drain to permit draining and cleaning.

The purpose of the waste outlet being trapped to retain 18 inches of water in the tank is so that the tank volume does not drop below 18 inches during a test cycle.  In practice what this means is that the flow from the grease interceptor can actually pass through the skim tank during a test cycle since the waste outlet is typically not closed off during testing.

Is this a problem?

Perhaps not, but since the point of the testing procedure is to determine with the greatest accuracy possible the efficiency of a grease interceptor, this doesn't seem right to me.

In case you didn't quite follow all that, allow me to clarify the issue.

The skim tank is just that a tank intended to capture the effluent from the grease interceptor during a test run in order to measure the amount of lard that escaped the interceptor which will determine the efficiency of the interceptor.

If the tank is open during a test run it is possible for some of the lard that enters the skim tank to escape the skim tank.  Any lard which escapes the skim tank is automatically counted as lard that must be in the grease interceptor, which has the potential of reporting a higher than actual efficiency of the interceptor for a given test cycle.

Since the interceptor will be tested over a minimum of 13 cycles, the problem can be compounded resulting in a potentially significant higher than actual average efficiency for a certified grease interceptor.

The solution would seem to be a simple one; keep the skim tank closed during each test cycle.

I have proposed just such a radical idea to the committee before but have received resistance from some who cite the benefit of the "automatic" nature of the testing when the tank is allowed to remain open during each test cycle.

I take the term "automatic" to basically mean, "easy-peasy", which is actually short for "easy-peasy-lemon-squeezy", which means "very easy" and makes me think about vodka and a drink called a Lemon Drop...

I doubt that the proponent of the argument intended committee members to conjure up images of
alcoholic drinks, but I did anyway.

Further arguments have been proposed that at higher flows and a two-drop series (allowed by the standard) that a lack of  human intervention at the required time would cause the skim tank to overflow, invalidating the test.

It's hard to not respond with a bit of sarcasm to that argument, but I'll abstain and simply say, "hogwash".

The committee has already agreed to enlarge the size of the skim tank for testing grease interceptors at flow rates over 50 gpm.  The size of the larger skim tank will allow for two complete drops (test increments) at 100 gpm, which is 400 total gallons of water. There is no risk of overflowing the skim tank unless the "human interaction" fails to drain the skim tank prior to another test cycle.

If testing personnel fail to drain the skim tank down between two successive two-drop tests at 100 gpm, then I would agree that the test results would be invalid, requiring the interceptor to be tested over from the beginning - and the testing personnel fired for incompetence (after they clean up the mess, of course).

Anyways, it seems to me that accuracy in testing should be the most important issue - definitely more important that what's easy-peasy.

I'll update this post if any progress on the issue is made at the next committee meeting.

No comments:

Post a Comment