Lacking the Scientific Method?
Updated: Mar 25, 2019
As a student of our collective fats, oils and grease (FOG) industry, naturally I follow relative studies, research projects, white papers and so on with great interest. I know many readers of this blog also try to stay abreast of the various investigations that might help strengthen FOG programs or provide more effective solutions to FOG abatement and pretreatment efforts.
In 2010 a study titled, Assessment of Internal and External Grease Interceptors for Removal of Food Based Fats, Oil and Grease from Food Service Establishments, by Gallimore E, Aziz TN, Movahed Z, Ducoste J, endeavored to compare the removal efficiency of a passive hydromechanical grease interceptor (HGI) and a automatic grease removal device (GRD) to a gravity grease interceptor (GGI) of a design approved by the Washington Suburban Sanitary Commission (WSSC).
The researchers concluded that the WSSC designed GGI consistently achieved the highest FOG removal under all of the tested conditions. The problem is that the conclusion may well be wrong! I understand that GGI proponents would naturally leap on any conclusion that supported the use of these devices over HGIs or GRDs, but must we cast all reason aside in the process?
Any honest evaluation of differing technologies ought to follow the scientific method if we are to find satisfactory conclusions available. One of the techniques that is essential in proving a hypothesis is to use a controlled experiment, meaning that only one (preferably) factor is changed at a time while all others are kept constant. Comparing multiple factors simultaneously makes it virtually impossible to distinguish which factor accounts for differences in outcomes between the various subjects of the experiment.
For example, as a golfer I am always interested in improving my game, as evasive as that goal appears to be. Lets say that three manufacturers all launched a brand new series of golf balls that boasted the longest distance ball on the planet. How would I go about proving which ball was actually the longest, assuming that it was not possible for them to all be the same?
Let's say that I took Ball A and went to the driving range and repeatedly hit it with my 9 iron and plotted the average distance at 145 yards. Then, on another day, I took Ball B and repeated the process using my 7 iron with a plotted average distance of 160 yards. Finally, I took Ball C on a different day and repeated the process using my 3 iron with a plotted average distance of 180 yards.
Which ball is the longest distance ball? If you are not familiar with the game of golf then let me clue you in that different numbered irons have different lofts which result in different distances depending on the club selected. Generally the lower the number of the club the lower the loft and hence, the longer the distance it will deliver a ball. How can we draw any conclusions about which ball is the longer distance ball if we used different clubs with different plotted average distances?
Wait, it gets more complicated. In my example I also went on different days. What was the weather like on each day? Was it raining? Was it windy? Was the wind blowing from behind or was it blowing in my face? Was I in Arizona on one day and Washington state on another day? Was it winter on one day and summer on another? Each of these factors will have different effects on distance. This is why any reliable comparison would require the elimination of as many variables as possible to simulate as closely as possible the same test conditions for one ball as for the others.
Herein lies the problem with the WSSC GGI study. There are variables between the HGI, GRD and GGI that are not accounted for in the conclusions drawn from the testing.
For example, the sizes of each of the grease interceptors are different: the HGI was 5 gallons, the GRD was 25 gallons and the GGI was 27 gallons. The flow rates that each grease interceptor was tested at were different: the HGI was tested at 10 GPM and 5 GPM, the GRD was tested at 25 GPM and 12.5 GPM and the GGI was tested at 0.9 GPM and 0.45 GPM. Also, each grease interceptor was tested at different retention times: the HGI was tested at 30 seconds and one-minute, the GRD was tested at one-minute and two-minutes, and the GGI was tested at 30 minutes and 60 minutes. Additionally, testing was conducted at three different emulsions; weak, medium and strong, and at two different temperatures (70 and 100 deg. F).
Regarding the emulsions, the paper makes the following admission, "Though the emulsion generation in each FGI [HGI] experiment was consistent, variations in flow and temperature caused differences in the globule sizes. In other words, the weak-emulsion condition for the 25gpm FGI [GRD] experiments produced smaller globules due to the shearing of the increased water flow compared to the weak-emulsion condition for the 10gpm FGI [GRD] experiments. A relatively stronger emulsion was produced for RGI testing than the emulsion produced for the FGI [HGI and GRD] testing, due to a higher emulsion generation flow needed for RGI [GGI] testing.”
It is impossible to determine the effect of all of these variables in the test conditions that were not accounted for in the test results. Any conclusions that might be drawn would be unreliable at the very least.
We all want to see field or bench testing that can corroborate performance claims and justify policies mandating specific solutions in FOG abatement efforts. Where jurisdictions should be cautious is in forming policy that is based on research that can be questioned because the process was somehow lacking in the Scientifc Method.