Wednesday, April 2, 2014

How do Gravity Grease Interceptors Work and How Efficient are they?

Many jurisdictions across the United States either favor exterior gravity interceptors or outright mandate them, some going as far as disallowing hydromechanical grease interceptors altogether.  

I get it.  I understand why.  Jurisdictions are frustrated!

For many decades grease traps (now called hydromechanical grease interceptors - HGI) have been the primary pretreatment device prescribed by plumbing codes to be installed in commercial food service establishments to prevent grease from entering the collection system.

If they work so well, why have jurisdictions been having so many issues with the buildup of FOG in collection systems?  

A simple inspection of the restaurants in a lot of jurisdictions will reveal a significant number of undersized indoor grease interceptors that are connected only to a multi-compartment sink and are not being maintained frequently enough.  

Hey, Americans are some of the best ditch divers (going from one extreme to another) in the world.  If the problem is allowing restaurants to install undersized grease interceptors inside the kitchen, then the "fix" is only allowing swimming pool sized gravity interceptors outside, right?!

How efficient are these giant interceptors?  How much grease will they hold?

I have asked jurisdictions across the country from sea to shining sea these same questions for the past 5 years, and you know what the answer always is?  "I don't know" - that's the answer. No one knows.  

How is it possible that we have jurisdictions mandating and in some cases only allowing the installation of interceptors whose efficiency and grease storage capacities are unknown?

I suppose the most common argument I hear is that basic physics justify their approval.  Okay, what physics are we talking about?  Stokes law.  The universally accepted answer as to why gravity interceptors work is because of Stokes law and retention time.

Really?  Okay, so what is Stokes law and how does it work?

Stokes law explained
The American Society of Plumbing Engineers (ASPE), Plumbing Engineering Design Handbook 4, Plumbing Components and Equipment, covers grease interceptors in chapter 8. Regarding Stokes law it has this to say on page 153: 

"An examination of this equation shows that the vertical velocity of a grease globule in water depends on the density and diameter of the globule, the density and viscosity of the water, and the temperature of the water and FOG material. Specifically, the grease globule's vertical velocity is highly dependent on the globule's diameter, with small globules rising much more slowly than larger ones. Thus, the larger the globule, the faster the rate of separation."

Table 8-1 Droplet Rise Time
Travel Time for 3" Distance at   68 deg F (hr:min:sec) Travel Time for 3" Distance at   68 deg F (hr:min:sec)
Droplet Diameter (microns) Oil (rise time) SG 0.85 Droplet Diameter (microns) Oil (rise time) SG 0.90
300 0:00:12 300 0:00:15
150 0:00:42 150 0:01:03
125 0:01:00 125 0:01:27
90 0:01:54 90 0:02:54
60 0:04:12 60 0:06:36
50 0:06:18 50 0:09:18
40 0:09:36 40 0:14:24
30 0:17:24 30 0:25:48
20 0:38:46 20 0:58:08
15 1:08:54 15 1:43:22
10 2:35:02 10 3:52:33
5 10:02:09 5 15:30:14
1 258:23:53 1 387:35:49

In the table above ASPE calculated the different rise rates of grease globules based on size in microns, and two different specific gravities (SG) - 0.85 and 0.90, showing how long it would take them to rise 3 inches at 68 deg. F.

Quoting ASPE again:

"Due to reliance on gravity differential phenomena, there is a practical limitation to interceptor effectiveness. In terms of grease/oil globule size, an interceptor will be effective over a globule diameter range having a lower limit of 0.015 centimeter (150 microns)."

Stokes law is fairly straight forward with one catch; it calculates the rate of rise of a grease globule in static (not moving) water. Does a grease interceptor, whether HGI or gravity, contain static water? Obviously not.  Either type of interceptor will have fixtures draining waste water into them at some flow rate.  How does this flow of waste water affect the calculation of Stokes law?

We need to understand the difference between laminar and turbulent flow before attempting to ascertain their affect on the calculation of Stokes law.

According to the Encyclopedia Britannica:
Turbulent Flow

  • Turbulent flow is fluid flow in which the fluid undergoes irregular fluctuations, or mixing. The speed of the fluid at a point is continuously undergoing changes in magnitude and direction, which results in swirling and eddying as the bulk of the fluid moves in a specific direction.
  • Laminar flow is fluid flow in which the fluid travels smoothly or in regular paths. The velocitypressure, and other flow properties at each point in the fluid remain constant. Laminar flow over a horizontal surface may be thought of as consisting of thin layers, all parallel to each other, that slide over each other. 
It is impossible to accurately predict the rise rate of a globule of grease in a turbulent flow environment because the turbulent flow acts to re-entrain the grease into the flow path negating the buoyant forces that are acting to lift the grease globule to the surface (depending on its size of course).

The advantage of a laminar flow environment is that the flow has very little effect (adding drag) on the buoyant forces acting on a grease globule.

Quoting again from page 155 of the same ASPE handbook:

"The ability of an interceptor to perform its primary function depends on a number of factors. These include the type and state of FOG in the waste flow, the characteristics of the carrier stream [turbulent vs. laminar], and the design and size of the unit."

Quoting from a paper published in 1944 titled Symposium on Grease Removal, Design and Operation of Grease Interceptors, by F.M. Dawson and A.A. Kalinske: 

"For simplicity let us assume that pure grease and water enter near the bottom of a rectangular-shaped interceptor L feet long, B feet wide, and with a water depth of D feet. The interceptor will do a good job of separation if, as the flow goes through the interceptor, the mean velocity of the flow is such as to permit a grease globule to rise a vertical distance D in a length of L feet."

Let's go back to our ASPE handbook:

"The ideal separation basin is one that has no turbulence, short-circuiting, or eddies.  The flow through the interceptor is laminar and distributed uniformly throughout the basin's cross-sectional area."

Distributing the flow uniformly throughout the basin's cross-sectional area reduces velocity which allows more time for a grease globule to separate inside the interceptor before reaching the outlet and escaping.  The velocity of the flow is reduced proportional to the extent to which it is uniformly distributed throughout the basin's cross-sectional area. 

The question is; do gravity interceptors have a laminar flow path and do they uniformly distribute the flow throughout their cross-sectional areas?
IAPMO/ANSI Z1001 Gravity Grease Interceptor

Here is what the standard gravity interceptor design looks like; the design that is mandated by IAPMO/ANSI Z1001 (formerly PS-80) and approved by model plumbing codes and most jurisdictions:

According to the Water Environment Research Foundations' 2008 report Assessment of Grease Interceptor Performance:  

Figure 5-18
"The standard configuration [IAPMO/ANSI Z1001] displayed substantially poorer performance than any of the other simulations (56.7%). Close observation of the velocity contours (Figure 5-18) of the cross-section within the reactor suggests that the cause for this reduced performance may be due to higher local velocities near the inlet, baffle, and outlet pipes, all contributing non-quiescent (turbulent) flow conditions...Further examination of the velocity contours in the direction of flow (Figure 5-19) suggests several regions of high velocity along the bottom of the tank and along the side walls."

WERF is essentially saying that at higher flow rates the lack of any control over the flow in gravity interceptors exacerbates turbulence and horizontal velocity leading to short-circuiting.

Figure 5-19
The simulations and bench tests pictured in Figure 5-18 and 5-19 represent a 20 minute retention time, which would be 50 gpm entering a 1000 gallon gravity interceptor. That is obviously a significant amount of flow and in fairness is higher than the average flow rate from most FSEs.

WERF's 2008 report also documented Grease Interceptor Influent Fluid Flow Analysis conducted on several different food service establishments (FSE) in Table 4-3 page 4-8.  The data collected included the total flow to the grease interceptor, the maximum flow-rate measured, the average flow-rate for the measurement period, and the size of the interceptor, among other things.

The outlier on maximum flow-rate across all restaurants was 45 GPM at a ‘full service steak house’. The average flow-rate at this same restaurant was 9.8 GPM.  The second highest flow-rate recorded was 35 GPM at a ‘full fare – Italian’ restaurant and the average flow-rate for this FSE was 9.4 GPM. The average flow rate for all restaurants was 2.8 gpm.

It is important to understand that the flow rates from these FSEs have low averages because restaurants don't typically fill up all their sinks and dump them repeatedly all day long.  Normal kitchen operations have some flow associated with both cooking and cleaning throughout the day, however, when they pull the drain plugs at the end of a meal period or the end of the day - this is when they will have a significant amount of flow at higher temperatures.

Based on all of the factors that effect grease interceptor performance, we would expect that a gravity grease interceptor would perform well at the average flow rates of most restaurants. However, we should also expect them to have problems with short circuiting at higher flow rates, and this is in fact what WERF reported finding in their analysis of interceptors in real world installations.  

On page 156 of the handbook we've been referring to here, ASPE explains:

"you can improve the grease interceptor by increasing the interceptor volume or reducing flow and subsequently lowering horizontal velocity and increasing retention time within the interceptor."

Therefore, gravity interceptors will actually need to be bigger in order to prevent short circuiting at higher flow rates - a decision that has consequences both positive and negative (i.e. better efficiencies but H2S gas generation).

Hydromechanical interceptors, though smaller in size and volume, have the advantage of being able to control the entering waste stream creating a laminar flow environment and distributing that flow more or less uniformly throughout their cross-sectional area, thereby reducing horizontal velocity and increasing flow through time.  This is why they are a viable alternative to larger gravity interceptors.

The challenge is to ensure that they are sized to have all of the kitchen fixtures routed to them, in order to avoid inadvertent bypass, and then to make sure they are maintained properly -(both subjects of another post).

1 comment:

  1. Nice blog about grease. The issues we face in CA are the city municipal utilities staff that don't understand how to accurately test the grease trap. Working with them & not against them has proved challenging - but working towards a common ground.