Grease Interceptor Testing for Miami-Dade FOG2.0
In late 2017 we were contacted by Mifab about testing 2 XL-MI-G-PL-1150 units in series in order to meet the requirements of Miami-Dade County. Section 5.0 of the FOG2.0 Control Device Guidance Manual, published by Miami-Dade County in May 2018, requires that units be tested in accordance with “ASME A112.14.3 Appendix A, ASME A112.14.4, CSA B481.3 or PDI G101.” and that “When proposing hydromechanical interceptors the engineer should indicate in plans … capacity in pounds of grease retention at 99% removal efficiency.”
The test was to be set up as ASME 112.14.3 Type A with a flow rate of 100 gpm. Mifab wanted to run until failure as define by 112.14.3/B481/G-101, which is a cumulative removal <90% or an incremental removal <80%.
The fact we were testing two units in series, loading solids in the first unit and targeting 99% removal made it obvious that this was not just another performance test. Exactly how different it would be really hit home when the skid containing 2000 lbs. of lard arrived at the lab. When another skid with 600 lbs. arrived a week later we hardly batted an eye. For more than a week we tested the performance of the units. The mountain of lard on the skids slowly decreased and the amount of lard retained in the skimming tank slowly went up.
The grease interceptor performance record book, if it exists, will need to be re-written. Never before has so much lard been added and retained in a test. Never before has a skimming tank been skimmed so many times. Never before have the words, “It is still working, seriously?” been uttered so often during a test.
A typical ASME test lasts 15 runs, this test lasted 8 times as long. For the first five runs, no lard made it to the receiving tank, 100% was captured. It was not until run #74 that the incremental removal dropped below 99%, to 98.9%. It was not until run #112 that the cumulative removal dropped below 99%. The Miami-Dade threshold had been crossed, at 2215 lbs. retained, but the unit had not yet failed the test. Even though we had predicted a capacity in excess of 2000 lbs it was still a bit of a shock to see the system keep performing. Finally, at run #125 the incremental efficiency dropped below 80% and the test was over.
The team breathed a collective sigh of relief, in part because we were not entirely sure it would ever end and in part because the performance met expectations. We now have a complete performance curve for two units in series. There is much more data than necessary but now we have a full data set that leaves no room for doubt. This will be a good baseline for planning future test protocols for large capacity systems.