2011 Progress Report: An Integrated Approach to Understanding and Reducing Fat, Oil, and Grease (FOG) Deposit Formation for Sustainable Sewer Collection SystemsEPA Grant Number: R834264
Title: An Integrated Approach to Understanding and Reducing Fat, Oil, and Grease (FOG) Deposit Formation for Sustainable Sewer Collection Systems
Investigators: Ducoste, Joel
Institution: North Carolina State University
EPA Project Officer: Page, Angela
Project Period: August 1, 2009 through July 31, 2012 (Extended to July 31, 2014)
Project Period Covered by this Report: August 16, 2010 through August 15,2011
Project Amount: $569,568
RFA: Innovative and Integrative Approaches for Advancing Public Health Protection Through Water Infrastructure Sustainability (2008) RFA Text | Recipients Lists
Research Category: Pollution Prevention/Sustainable Development , Sustainability , Water
The objectives of this study are as follows: 1) perform bench scale experiments that attempts to recreate FOG deposits and determine parameters that significantly influence their formation rate, 2) develop a numerical model that describes FOG deposit formation kinetics, 3) perform bench scale tests to explore enhanced treatment methods to improve the removal of FOG deposit chemical precursors with grease interceptors, 4) perform pilot scale experiments on a continuous flow sewer collection system to explore spatial variations in FOG deposit formation, and 5) develop a modified EPA storm water management model (SWMM) to predict FOG deposit formation in sewer collection systems.
Progress has been made in the second year for most of the project tasks including 1, 3, 5, 6, 7, 8, and 9. Overall progress on the project continues to be slowed due to significant illness of one of the hired graduate students and the Fourier Transform Infrared (FTIR) instrument being down for 5 months. We have upgraded an additional FTIR instrument so that the lab can utilize two units for the project. Hopefully, the additional unit will provide more opportunity to perform all the measurements at a faster pace. Measurement and analysis of the data require more time than anticipated. The research team continues to try to make up for lost time.
One of the underlying assumptions in this research project has been that FOG deposits are fatty acid salts of calcium or calcium-based soaps. While there were overwhelming clues to this assumption from the data (FFA profile, metals and mineral analysis, structural tests) collected by the PI on a previous research project and published (Keener, et al., 2008), it had remained a hypothesis because only the ingredients were confirmed and not the presence of actual compound bonds. Recently, the PI invested some resources using the FTIR instrument and a graduate student working on this project to test this hypothesis. FOG deposits created under laboratory conditions with the addition of calcium in a solution containing grease interceptor effluent waste from a food service establishment were analyzed and compared with fatty acid salts of calcium created with oil and calcium under ideal alkali hydrolysis conditions. Results of the analysis of both solids revealed that a saponification reaction did occur and that the FTIR spectra were similar for both solid samples. FTIR analysis also was performed for actual FOG deposit samples collected from a local sewer collection system. The results of these tests also confirmed that the FOG deposits exhibited bonds that were consistent with a saponification reaction.
An important observation, while computing percent saponification of calcium soaps made with calcium sulfate, was that calcium sulfate both in its pure form and in solution with DI water contains metal oxygen (Ca=O) bond. This observation led to difficulty in ascertaining the contribution of Ca=O bond (i.e., from calcium sulfate itself, or from saponification) in the determination of the percent soap for calcium sulfate based soaps. Calcium sulfate in its different forms (anhydrite, mono-hydrate, di-hydrate) also exhibit distinctly different solubility, which affects the available amount of Ca2+ ions in the solution to take part in the saponification reaction. The mixing process followed in this study also caused a noticeable difference in creating a uniform mixture of calcium sulfate solution and liquid/melted solid fat. Calcium chloride, having a higher solubility than calcium sulfate, produced a uniform mixture and hence a consistent mass.
A modification in the addition process of calcium source was performed in the experiments with beef tallow under high temperature conditions. Beef tallow was melted first at 45°C. The high mixing condition was used to make the melting process faster after a significant amount of beef tallow already had been melted. Calcium source then was added to melted fat at the high intensity region (around the tip of the mixing blades) with the mixing blades already working inside the reactor. A dramatic difference in the reactor mixture was observed in the aforementioned process of adding calcium sulfate solution to canola or beef tallow under high temperature condition. No separation behavior was observed and a uniform mixture appeared with both fat types. A possible interpretation is that slow initial contact between liquid/melted fat and calcium sulfate solution imparted properties similar to tackiness of the calcium sulfate mass, which later led to the coagulation of small and separate calcium sulfate masses. The interpretation is providing the research team with reasonable insights into the FOG deposit accumulation process occurring in the high calcium environment in the sewer crown space as a result of the potential calcium leaching out of concrete.
The results of the saponification reaction kinetics involving calcium chloride show that an average 15-20 percent of soap was formed within minutes of mixing reactant constituents, due to instantaneous alkali driven hydrolysis, for all temperature and pH conditions. The instantaneous amount formed in case of calcium sulfate based soaps was found to be an average of 60-65 percent. However, as previously mentioned, for calcium sulfate based soaps, a major portion of initial soap computation could be due to the initial presence of the Ca=O bond in calcium sulfate.
The soap formations for calcium chloride based soaps were observed to decrease by 20 percent with increasing temperature due to the decreasing solubility of CaCl. The soap formations for calcium sulfate based soaps after 8 hours of reaction time were observed to remain the same (60 percent average). However, with increasing reaction time, the contribution of Ca=O bond from the soap was assumed to increase. In-depth analysis is underway to better understand the contribution of metal oxygen bonds in earlier vs. later stages of saponification reaction.
Significant amount of solid/semi-solid soaps were observed to form in all experiments. Rheological results of the final soap samples indicate that moist soaps exhibit elastic behavior more than viscous in the linear visco-elastic region. The calcium soaps, have a gel structure exhibiting a certain form of stability. A gel point was observed indicating a transition from viscous to elastic properties with additional application of shear. However, the interpretations of the rheological results are still under exploration. More analyses are yet to be performed for better understanding of the rheological data.
Empirical kinetic models developed by Cotte, et al. (2006) and Foubert, et al. (2002) are being tested to determine their ability to predict the experimental soap kinetics data. Preliminary results of model predictions show that the Cotte, et al. (2006) model can reasonably represent saponification data. However, some experimental results display significant soap formation in the later stages of the test period. This jump in soap formation is hypothesized to be caused by latent autocatalytic processes that result in a higher saponification reaction rates. Because the current empirical models are unable to capture this later event, a mechanistic model for alkali driven hydrolysis based on mass action principles has been proposed. The model is based on sequential hydrolysis of triglyceride to the formation of intermediates (diglyceride and monoglyceride) that sequentially release free fatty acids (FFA). FFA then is available to react with calcium (Ca2+) to form soap. This new mechanistic model will be used to simulate the formation of FOG deposits in the pilot- and full-scale sewer collection systems and validated with the bench-scale soap formation kinetic data.
The pilot has been constructed and is in operation. The pilot sewer system consists of a plain pipe section, root intrusion pipe section, section containing a simulated pipe sag, and manholes. The pilot system has been in operation for 1 month to test the behavior of the injection system and the development of any solids buildup at different points along the pipe network. Based on 1 month of operation, the system displays the formation of solids in all the manholes. There also has been the formation of solids at the ridges of the fittings, which serve as the entry and exit of the manholes. A significant amount of solids formation also has been observed on the root intrusions portion of the sewer system. The solids in this section seem to be occurring directly upstream of the root system with little formation behind it in the straight section of pipe before entry into Manhole 2. The maximum amount of solids deposition was observed in the section of the pilot system surrounding the pipe sag bend. The solids buildup around the pipe sag seems to be in response to the low flow condition that is unable to force the accumulation of oil and potential reaction with calcium near the top portion of the pipe sag. None of these solids have been analyzed yet with the Fourier Transform Infrared-Attenuated Total Reflection (FTIR-ATR) instrument. The instrument is being repaired and should be operational soon. The FTIR-ATR analysis will provide definitive proof of whether the solids buildup that has been observed in the pilot system is actually calcium-based soap. Samples will be scraped from different sections of the system and categorized accordingly.
The research team has attempted to modify the EPA SWMM code to include the transport of reactive species through the sewer collection system. Models proposed and developed in this research project (Task 3) were difficult to include in the current framework of EPA SWMM without significant changes to the model structure. Consequently, the research team pursued an alternate strategy that involved a simulation framework called CITYDRAIN (Achleitner, et al., 2007) based on MATLAB Simulink code and is available without any commercial license other than MATLAB. CITYDRAIN utilizes the block diagram format of SIMULINK to represent different parts of the sewer system including catchments, sewer pipes, storage units, and receiving waters as well as the inclusion of a wastewater treatment plant.
CITYDRAIN was modified to include the proposed alkali saponification reaction model described in Task 3. The model includes tracking such constituents as the three forms of FOG (triglyceride, diglyceride, and monoglyceride), free fatty acid, calcium, and calcium-based fatty acid salt (i.e., FOG deposit) that occurs within the flowing wastewater as well as the portion that partitions to a solid surface. Preliminary model results predict buildup of FOG deposits in the three pipe sections of the pilot system with the largest buildup occurring in the pipe containing root intrusion. None of the manhole structures have been simulated in the preliminary model.
Keener, K. M.; Ducoste, J. J.; Holt, L. M. Properties influencing fat, oil and grease deposit formation. Water Environ. Res. 2008, 80, 2241–2246.
Cotte M., Checroun, E., Susini, J., Dumas, P., Tchoreloff, P., Besnard, M., Walter, P., 2006, Kinetics of Oil Saponification by Lead Salts in Ancient Preparations of Pharmaceutical Lead Plasters and Painting Lead Mediums, Talanta, 70, pp. 1136-1142.
Foubert, I., Vanrolleghem, P.A., Vanhoutte, B., Dewettinck, K., 2002, Dynamic Mathematical Model of the Crystallization Kinetics of Fats, Food Research International, 35, pp. 945-956.