I have a Bone to Pick with GGI Sizing

April 25, 2016

On April 18-19 the Western States Alliance hosted their first "official" FOG Forum in beautiful Bend Oregon. I was asked to give a presentation on how to properly size and select grease interceptors. During the presentation I made the comment that the 30 minute retention time has no basis in scientific data or research that would justify its use for sizing GGIs.

I was in the middle of explaining that the origin of the 30-minute retention time was uncertain but that the justification for its use was founded upon a textbook titled, Small and Decentralized Wastewater Management Systems. published in 1998 by McGraw-Hill. The authors of that textbook provided the following technical argument for a minimum 30-minute retention time when using a septic tank as a commercial grease interceptor, "Typically, skimming or interceptor tanks are used to trap oils by flotation and grease by cooling and flotation. The contents of the tank serve as a heat exchanger cooling the incoming liquid, which helps to solidify the greases. For flotation to be effective, the interceptor tank must detain the fluid for an adequate period of time (typically greater than 30 minutes)."

During this portion of the presentation an audience member said, "actually the 30 minute retention time was first introduced in the Uniform Plumbing Code in 1982 in Appendix H." Normally I can quickly ascertain what is wrong with a statement like that but honestly, at the time by brain cramped on me. So naturally I spent the next several days meditating on what I should have said. 

Since I should have had an answer to that statement but didn't, I decided to go back and review the Appendix H sizing method and to confirm what I already knew; that it was not based on a 30 minute retention time.

Here is the exact section from the 1982 UPC Appendix H:

Notice that there is a calculation for retention time and that it does not use 30-minutes. Instead notice that there are two options; for commercial kitchen wastes with dishwasher you use 2.4 hours (144 minutes) or for single service kitchens use 1.5 hours (90 minutes). Of course the time increases with storage factors added into the calculation. 

Okay, so the question remains; where did the 30-minute retention time come from?

Prior to the 1982 edition of the UPC and the then new Appendix H, the EPA published a Design Manual titled, Onsite Treatment of Wastewater Disposal Systems (1980). The manual provided the following formula for sizing GGIs:

The sizing formula provided in the EPA manual is similar to the Appendix H method but obviously not identical. Both though, lead to very, very large GGIs with retention times well in excess of 30-minutes. 

IAPMO formed a task group (TG) on grease interceptors to make recommendations for updating the requirements for these devices in the 2006 edition of the UPC. Among other things, the TG recommended replacing the Appendix H sizing method because, "No sound technical basis could be found to justify retaining the meals per peak hour sizing criteria found in Table H-1 (in Appendix H)." Instead the TG recommended a Drainage Fixture Unit (DFU) sizing methodology for GGIs that was based on a 30-minute retention time as supported by the very textbook from McGraw-Hill I noted at the beginning of this post. In other words the basis for using a 30-minute retention time was and still is anecdotal.

I have not found any evidence to support a technical or scientific basis that justifies using a 30-minute retention time in sizing GGIs. Anecdotal evidence cannot be the basis of any sizing methodology. The fact that these devices are not tested and rated for performance is the very reason that sizing is still an issue that I, for one, consider unresolved. That's why I have a bone to pick with GGI sizing.


  • Greg Williams

    May 03, 2016

    I do not own a boat so I will have to wade into this discussion. I like Max’s point about trying to take an engineering based approach to modelling interceptor behaviour. I also like Ken’s point that it is not so simple. In fact, I will see Ken’s “it’s complicated” and raise him a “fiendishly”.

    In order to calculate particle velocities you need to calculate forces. The drag force will be working horizontally and vertically against the buoyancy force, and gravity will be working against it vertically. They all depend on particle size in some way or another and particle size is not well known. One you have a velocity you need a distance to determine if the time available is enough. Particles do not rise straight up, so what is the distance? You can calculate streamlines, but what size particle ends up on which streamline?

    Modelling this situation, even with computational fluid dynamics, would be fiendishly difficult. The presence of turbulence and the physical size of a typical unit do not help. Under the circumstances, as Ken suggests, testing is the only viable option. The problem will be, if the testing protocol is bound more by the existing “sink dump” test than by the known and measurable law of physics then it will be difficult to make progress.

  • Ken Loucks

    May 03, 2016

    I’m not sure what Max means by a “proprietary red herring”. I don’t even own a red herring, or any other herrings for that matter. I have fished for Spring Chinooks using herrings, but they were not mine either (the herrings I mean).

    There is no dispute that gravity-differential separation, based on Stokes law, is the basis for the operation of any grease interceptor.

    The problem that I have with the 30-minute retention time, or even the “meals per peak hour” from Appendix H, is that neither has a defensible technical or scientific justification.

    The argument put forth justifying the 30-minute retention time appears to be supported by a theory that Stokes law explains the operational efficiency of GGIs. The problem with this theory is that Stokes law depends upon a laminar (turbulent free) flow environment in order to make accurate predictions of rise rates within the interceptor. The WERF report previously cited identified turbulence inside of GGIs as a primary cause of short-circuiting. As flow rates inside of GGIs increases the turbulence increases nullifying any Stokes law calculations.

    The theory put forth by Max argues that the law is both established and well known in fluid dynamics and thus the law is beyond question as the basis of operation in GGIs. The law is not in question, it’s the assumption that the law applies uniformly to GGIs that is theoretical and unsupported. If it can be proven that the flow environment in a GGI is turbulent free then it may be possible to prove that Stokes law is the primary predictor of performance. The WERF report has already proven that this is not the case.

    The fact that GGIs are not tested and rated for performance is the very reason that a “compromise” was required in order to select the 30-minute retention time for sizing GGIs. Unfortunately the technical support cited for using a 30-minute retention time turns out to be anecdotal. Which means neither the Appendix H method nor the 30-minute retention time are technically based sizing methodologies for GGIs.

    We ought to stop guessing and speculating on how GGIs work or how efficient they are. Instead we ought to insist upon a testing and rating protocol that provides predictable and repeatable results based on known and measurable laws of physics to establish the performance of any grease interceptor, including GGIs.

    Now, if you plan to go fishing for Spring Chinooks, herrings work well, but don’t forget some roe or even shrimp! If you’re feeling lazy you can always drag a plug, which works once in a while especially if you wrap the plug in some herring.

  • Ken Loucks

    May 03, 2016

    The “D” in the EPAs 1980 manual is for the number of seats in the Dining area.

  • Max Weiss

    May 02, 2016

    This is a proprietary red herring. There is no mystery why 30 minutes design retention time for gravity [Stokes] separation was settled upon as the most reasonable compromise between the optimum (longer retention time) and the expedient (shorter retention time).

    The question was addressed and decided by API [American Petroleum Institute] in the 1950’s. [API, 1969, Manual on Disposal of Refining Wastes; American Petroleum Institute, Division of Refining: New York, NY]

    The answer has been computationally available since 1851 when the “mathematical equation expressing the vertical velocities of small spherical particles in a fluid medium was first set forth by the British scientist Sir George G. Stokes it is derived by consideration of the forces acting on a particular particle as it sinks or rises through a liquid column under the influence of gravity.” And, the retention time topic is published in many fluid mechanics publications.

    Simply put: when horizontal velocity (flow) of a FOG particle exceeds vertical velocity (ascension) it passes through a Stokes -Gravity Grease Interceptor. The size of FOG particles, optimally 150 microns or larger, significantly affects ascension rate. The rate is determinable by completing Stokes formula with known values. Studies have shown FOG influent particle size distribution between 80 -120 microns is statistically most prevalent [WERF].

    The WERF study published comparisons of the effect of retention time on separation efficiency in a conventionally piping configuration Gravity Grease Interceptor, significantly summarized here: Standard Configuration (20min) = 56.7% removal; Standard Configuration (1hr) = 97.5% removal; 30 minutes = 76.4%. The data coincides with numerous studies documenting removal efficiencies of 70% – 80%.

    The physics of gravity separation of dissimilar density fluids is a long ago, a well developed field of science. The illuminating new area of interceptor investigation is the chemical environment and its effects on collection systems.

  • the Doctor

    Apr 29, 2016

    Ken, i like your post and have a question that maybe i should know. What is the “D” stand for on the calculation from “Onsite Treatment of Wastewater Disposal Systems (1980).”

Leave a comment

Please note, comments must be approved before they are published