BioCycle January 2008, Vol. 49, No. 1, p. 53
Results from full- and bench-scale studies suggest utility district’s food waste recycling process is suitable for cleaning and recycling postconsumer presorted solid food wastes while increasing biogas production in biosolids digesters.
Donald M.D. Gray (Gabb), Paul J. Suto and Mark H. Chien

WITH food residuals comprising significant tonnages in the municipal solid waste stream, many government agencies are targeting those organics for diversion. Anaerobic digestion could be an ideal solution for recycling food wastes into renewable energy (methane) and fertilizer, but food wastes from restaurants, grocery stores and other food handling facilities (postconsumer) can be highly contaminated, even when source separated, and are too dry (25 to 35 percent total solids) for wastewater treatment plant anaerobic digesters.
The East Bay Municipal Utility District (EBMUD), in Oakland, California, had excess digester capacity at its wastewater treatment plant because many food processors that were major contributors to the wastewater generated in the EBMUD service area have left the San Francisco Bay Area. To make use of this excess capacity, EBMUD contracted with a local hauler (Norcal Waste Systems) to bring source separated food waste to EBMUD digesters. The hauler pretreated the food waste using a trommel screen, a magnet for metals removal, and a hammermill for material size reduction and contaminant removal, steps typically used to pretreat solid wastes for composting and other waste recycling processes. EBMUD built a below-ground slurry tank to accept and dilute the food wastes from the hauler’s trucks. The first load of postconsumer solid food waste was received in May 2004.
EBMUD has continued to receive and anaerobically digest up to 40 tons/day of postconsumer solid food wastes ever since. Unfortunately, the hauler’s cleaning system did not provide a waste that was trouble-free for EBMUD’s treatment process. Grit, mostly from ground seafood shells, accumulated into piles in the slurry tank and plugged the tank outlet pipe, requiring frequent cleanings with a vacuum truck. Progressive cavity (PC) pumps that transferred the slurried food waste to the digesters were badly damaged when metal pieces from knives and forks, coins, nuts and bolts, and other contaminants in the food waste became embedded into the pumps’s stators, destroying the stators and scoring the pumps’s rotors. There was also concern about sending other nonbiodegradable materials such as plastics and fibrous materials (e.g., mop heads and rags) to the digester.
CONTAMINANT SEPARATION
To separate out these contaminants, process equipment such as vibrating screens and vortex separators were tested. These systems were unsuccessful, mostly due to blinding and plugging from the fibrous materials in the food waste slurry. Then, EBMUD staff developed a process that successfully separated the contaminants from the digestible food wastes. The process (patent pending and shown in Figure 1) includes the slurry tank, but upgrades the mixers based on computational fluid dynamics (CFD) analyses to reduce grit accumulation in the tanks. Next, a rock trap/grinder removes the heavy materials, such as metal pieces and rocks, from the slurry and grinds the bigger solids and fibers making the slurry more consistent and less prone to causing line plugging or negatively impacting paddle finisher performance. The slurry is then pumped, with a peristaltic pump designed to handle abrasive copper slurries in the mining industry, to a paddle finisher – equipment typically used in the food industry to remove stems, seeds and skins from fruits and vegetables to make juices, sauces, jams, etc.
The paddle finisher is composed of two to four paddles that rotate along the inside length of a cylindrical screen, with 0.045 or 0.060 inch openings; modifications to the size openings could be explored. Soft biodegradable materials are pushed and extruded through the screen, and are referred to as “pulp.” The harder, tougher and fibrous materials, larger than the screen openings (called “pomace”), are transported by the paddles down the inside length of the screen and pushed out of the finisher. The pulp is pumped to EBMUD’s anaerobic digesters; the pomace is trucked to a landfill, but eventually could be recycled with a dry digester (producing more methane gas for electricity generation), gasified (producing a synthetic gas for electricity generation and less residual solids) or recycled using some other process. The EBMUD system requires that all food waste be small enough so that the slurry can be pumped through the process; therefore, some form of material size reduction prior to processing may be necessary.
This article presents operational data from the EBMUD food waste recycling process, and results from a bench-scale study conducted by EBMUD staff and funded with a U.S. Environmental Protection Agency (EPA) grant (see sidebar). The bench-scale anaerobic digestion studies were done at both mesophilic and thermophilic temperatures to determine biodegradability, methane gas produced, and the minimum mean cell residence time (MCRT) needed for digestion when food waste pulp produced from the EBMUD food waste recycling process is fed to anaerobic digesters.
BENCH-SCALE METHODOLOGY
The EBMUD food waste recycling facility was operated at full-scale, accepting up to 40 tons of food waste/day. Processing flows were up to about 250 gallons/minute (gpm). Currently, the digesters receive pulp from EBMUD’s food waste recycling process. The slurry/pulp is pumped into a common pipe that distributes it among all of EBMUD’s in-service digesters (usually 6 to 7), along with other wastes being fed, primarily municipal wastewater solids. Digesters are each about 1.7 million gallons. The District is pursuing the possibility of having one or two digesters dedicated to food waste only, in order to produce a certified organic soil amendment or fertilizer (which isn’t possible if biosolids are also in the digestate).
For the bench-scale study, two 30-L anaerobic digesters were used. One operated at mesophilic temperature (meso), approximately 35°C, and the other operated at thermophilic temperature (thermo), 50 to 52°C. The meso digester was started using about 27 L of digested sludge from EBMUD’s full-scale mesophilic digesters, and the thermo digester was started with about 27 L of digested sludge from EBMUD’s full-scale thermophilic digesters.
Throughout the study, both digesters were fed 100 percent postconsumer food waste from EBMUD’s food waste pretreatment system (food waste pulp). Digested sludge in both digesters was monitored continuously with thermocouples connected to LabVIEW® Virtual Benchlogger™ software run on a laptop computer. Digested sludge pH was measured in situ to avoid inaccuracies due to carbon dioxide loss occurring from samples removed from the digester. Digested gas flow was measured with wet-tip gas meters (Wet Tip Gas Meter Co., Nashville, Tennessee). The digesters were operated at 15-, 10- and 5-day mean cell residence times (MCRTs), for at least 2 MCRTs of continuous and stable operation at each residence time (e.g. total of 30 days for a 15-day MCRT). The intent was to learn more about how postconsumer food waste pretreated with the EBMUD process is anaerobically digested, by measuring process performance at different MCRTs. This is significant from a statistical viewpoint since theoretically after 3 MCRTs (or 3 standard deviations from the mean), digester contents have been completely replaced with new microorganisms, waste, digester sludge, etc. during the test, and thus the responses are purely from the conditions of the test, and not influenced by the previous conditions. (For example, going from a 15-day to a 10-day MCRT.) In other words, the digester is completely in equilibrium with the current test conditions.
RESULTS
From analyses run on the EBMUD food waste recycling process, the pomace rejected was only about 10 percent or less of the total solids and only about 5 percent of the total chemical oxygen demand (COD) in the food waste slurry fed to the paddle finisher (the rest being in the pulp); total solids was about 25 to 31 percent. The pulp was 4.5 to 16 percent total solids, with 87 to 95 percent volatile, a biochemical oxygen demand (BOD) between 32,000 and 100,000 mg/L, and a COD between 85,000 and 222,000 mg/L.
The paddle finisher not only removes nonbiodegradable materials from the food waste slurry, it also reduces particle size by extruding softer materials through its fine screen (Figure 2). Sanders et. al. (2003) showed that the hydrolysis rate constant increases almost three times when the organic particle size is reduced from 2,000µ down to 60µ, suggesting that the pulp should biodegrade faster in the digester than the original waste slurry would have degraded.
Volatile Solids Destruction: Results from the bench-scale study suggest that food waste pulp from the EBMUD process is highly biodegradable. Figure 3 shows the high volatile solids destruction (VSD) of the pulp in the thermophilic digester. Often the pulp’s VSD was 80 percent or higher, even at a 10-day MCRT, compared to only about 55 percent VSD for municipal sludge at a 15-day MCRT (WEF, 1998). The volatile portion of the pulp’s total solids was about 85 to 90 percent (Figure 4), compared to only 77 percent volatile solids for municipal sludges. Based on these numbers, the mass of biosolids produced from anaerobically digesting food waste pulp would be about half or less than that produced from anaerobically digesting municipal sludge.
Volatile Solids Loading: Figure 5 shows the high volatile solids loading (VSL) rates the bench-scale digesters were able to accept when food waste pulp was fed. During the 10-day MCRT period, the food waste pulp VSL was about three times as high as the recommended maximum loading rate of 0.20 lbs VS/ft3-day for anaerobic digestion of municipal sludges (WEF, 1998). This maximum loading rate was based on the accumulation rate of ammonia and other toxics (WEF, 1998), suggesting that the food waste pulp contains lower concentrations of toxic materials than municipal sludges. Figure 6 shows the much higher COD loading rate possible with food waste pulp (approximately 1.25 lbs COD/ft3-day) compared to municipal sludges (0.06-0.3 lbs COD/ft3-day, per Metcalf and Eddy, 2003).
Total Solids: Figure 7 shows the average food waste pulp total solids (TS) fed to the bench-scale digesters was about 10 percent during much of the 10-day MCRT period – clearly above the average 4 percent TS for municipal sludges fed to anaerobic digesters by more than two times. The digested sludge TS appears to rise with the higher feed TS for both meso and thermo digesters, with the meso jumping up to 4 percent TS, and the thermo digester as high as 3 percent TS. Nonetheless, digested sludge from both digesters eventually returned down to 2 percent TS, consistent with digesters fed municipal sludges at 4 percent TS. This suggests that food wastes pretreated with the EBMUD process might safely be fed at concentrations even higher than 10 percent, after sufficient acclimation, without interfering with digester mixing efficiency, which could significantly increase digestion capacity in existing digesters. Since food waste is received at 25 to 30 percent TS, this only means that less water would need to be added to boost the feed solids higher than 10 percent.
Mean Residence Time: The data presented so far appears to show that food waste pulp could be digested adequately at MCRTs as low as 5 days. Other data, however, suggest that a 10-day MCRT is adequate, but a 5-day MCRT is too low. Figure 8 shows a rapid drop in digested sludge pH, down to about 5.5, for both meso and thermo digesters after their MCRTs were lowered to 5 days. The figure also shows that 15- and 10-day MCRT periods had pH drops as well. These were mostly due to heater failures when digester temperatures dropped quickly and dramatically. In these cases, the digesters were restarted with fresh digested sludge and feeding was resumed. Consistent and stable operation was maintained for at least 2 MCRTs (15- and 10-day) in both meso and thermo digesters. The bench-scale digesters, both meso and thermo, were operated at a 5-day MCRT continuously and at consistent temperature for 3 MCRTs.
Digester Gas Production: Figure 9 shows a rapid drop in digester gas production after the MCRT was lowered to 5 days. Figure 10 shows a very sharp decline in digester gas methane content in both meso and thermo digesters when operated at a 5-day MCRT. These two figures suggest that methanogens have been severely impacted in both meso and thermo digesters when the digester MCRT was reduced to 5 days. Based on data presented previously, it is more likely due to methanogen populations being diluted out of the digester’s sludge at the lower MCRT, because their growth rates are too slow to keep up with the higher digested sludge withdrawal rates, rather than organic loading rates being too high.
Figures 9 and 10 also show the significantly higher methane production from food waste pulp compared to the municipal wastewater treatment solids, when anaerobic digesters are operated at a 15- or 10-day MCRT. This demonstrates that approximately 3 to 3.5 times as much methane can be produced per unit of digester volume from food waste pulp than from municipal wastewater solids (Table 1). Because of this, and the lower MCRT required for food waste pulp, a smaller digester and lower capital costs appear possible for food waste pulp compared to digesting municipal wastewater solids.
Figure 11 most clearly shows process failure at the 5-day MCRT, with relatively steady digester sludge alkalinities during the 15- and 10-day MCRT periods, but the very sharp alkalinity decline directly after lowering the MCRT to 5 days. Table 1 summarizes the comparison of food waste pulp to municipal sludges found in this study.
Thermophilic Vs. Mesophilic: Both meso and thermo bench-scale digesters adequately digested the food waste pulp at a 15- and 10-day MCRT with similar results. Nonetheless, the thermo digester appeared to perform slightly better when considering a few of the digester’s performance parameters. Digested sludge total solids (TS) and volatile solids (VS)/TS ratio was lower for the thermo than for the meso digester, indicating that the thermo digester may have been able to better biodegrade the pulp than the meso digester (see Figures 7 and 4). The thermo digester also appeared to have consistently higher volatile solids destruction (VSD) of the pulp than the meso digester (Figure 3). The thermo digester also may have produced slightly more digester gas (Figure 9) and with a slightly higher methane content (Figure 10) than the meso digester. More study may be needed in this area to determine if there is a significant performance difference between the two operating temperatures (this study suggests that there may be).
DISCUSSION
Results from this study show that both mesophilic and thermophilic digesters fed only the food waste pulp have volatile solids reductions between 75 and 85 percent at a 15-day MCRT. With volatile solids at 87 to 95 percent, this demonstrates how very biodegradable the food waste pulp is, compared to municipal sludges, in an anaerobic digester. The results also suggest that anaerobically digesting a 10 percent TS food waste pulp at a 10-day MCRT is practical, and that digesting even higher pulp concentrations may be reasonable.
Operating at higher pulp concentrations and lower digester MCRTs could significantly reduce the digester volume needed. For example, consider two cases: 1) A 10 percent pulp digested at a 15-day MCRT, and 2) A 15 percent pulp digested at a 10-day MCRT. If the 20-ton food waste loads arrive at 30 percent TS, a 10 percent pulp would require a 14,400-gallon slurry and a 15 percent pulp a 9,600-gallon slurry. Assuming 100 tons of food waste/day, 5 days/week, or 71.4 tons/day average, Case 1 would require a 770,000-gallon digester, while Case 2 would only require a 343,000-gallon digester, or only about 45 percent of the digester volume of Case 1. So, at Case 2’s higher pulp concentration and lower MCRT, half the digester volume, or half the number of equally-sized digesters, would be needed to process a given amount of food waste pulp than at Case 1’s higher MCRT and lower pulp concentration.
EBMUD has been taking food waste to its full-scale anaerobic digesters for over three years, and has identified many problems and solutions with taking this type of waste to wastewater treatment plant anaerobic digesters. EBMUD also has demonstrated the viability and benefits of taking food waste to a wastewater treatment plant anaerobic digester, e.g., digesting 40 tons/day of food waste (two truck loads taken by EBMUD a day) is estimated to produce enough renewable energy to power over 600 households. Compared to municipal sludges (wastewater treatment plant primary and secondary sludges), this study suggests that food waste pulp produces more methane, but requires less than half the digester volume, and produces less than half the biosolids per pound of food waste pulp fed than municipal sludges. EBMUD is considering increasing food waste acceptance up to 200 to 300 tons/day, which could produce power equivalent to 3,000 to 4,500 households or more. Using EBMUD’s experience, other communities could turn their food waste problem into green energy at their own wastewater treatment plant.
The food waste treatment process developed by EBMUD has been demonstrated to process 20 tons of food waste in about one hour with one paddle finisher, removing nonbiodegradable contaminants greater than 0.040 inches and producing a pulp containing about 95 percent of the COD from the original load – providing a highly bio-degradable pulp material without the contaminants found in food wastes that can impede or harm wastewater treatment plant processes. With further operational experience, equipment maintenance requirements and service life for all the process’s components will need to be assessed, since their use in this type of process is unique.
The EBMUD process also might be used for other purposes such as cleaning municipal sludges (wastewater primary and secondary sludges) prior to anaerobic digestion to maintain a larger digester active volume and reduce the need to clean digesters, to produce a higher quality biosolids and to possibly increase volatile solids destruction and reduce biosolids production. Fats, oils and greases
(FOG) also might be pretreated in this process prior to anaerobic digestion, especially if a heating system to reduce grease accumulation on the finisher screen could be implemented. Even digested sludge could be recycled through the EBMUD process to possibly reduce digester downtime for cleaning.
STUDY CONCLUSIONS
The following conclusions were based on bench-scale anaerobic digestion of food waste pulp produced by the EBMUD food waste recycling process:
o Minimum MCRT required to anaerobically digest food waste pulp is 10 days or less, but is greater than 5 days.
o Anaerobic digestion appears to be able to handle much higher volatile solids and COD loading rates from food waste pulp than from municipal sludges, possibly by as much as three or more times higher.
o Food waste pulp fed to anaerobic digesters at 10 percent TS appears to produce a 2 percent digested sludge, similar to feeding a 4 percent TS municipal sludge. This suggests that food waste pulp might be fed at higher TS concentrations than 10 percent without impacting digester mixing.
o From the items above, the anaerobic digester volume needed for food waste pulp might be half or less than the volume needed to digest municipal sludges.
o Food waste pulp appears to be more biodegradable than municipal sludge because of its higher volatile solids portion (85 to 90 percent compared to 70 to 80 percent for municipal sludge) and its higher volatile solids destruction (80 percent compared to only 56 percent for municipal sludge) under anaerobic digestion.
o Solids production from food waste pulp appears to be half or less than that produced from municipal sludge.
o Methane production from food waste pulp appears to be as high or higher than that produced from municipal sludges. For the same amount of methane produced, results from this study suggest that about half or less digester volume is needed and about half or less biosolids are produced.
o Although food waste needs cleaning prior to anaerobic digestion, and this has a cost, overall food waste pulp might produce methane at a lower cost than municipal sludge does. More investigation, however, is needed to better compare methane production costs from the two feedstocks.
o Anaerobic digesters appear to perform slightly better at thermophilic compared to mesophilic operating temperatures when fed food waste pulp, based on this study’s volatile solids destruction, methane production and other results; but further investigation is needed to confirm this.
ACKNOWLEDGEMENTS
The bench-scale anaerobic digestion study was funded by a U.S. Environmental Protection Agency grant No. EPA-R9-WST-06-004. The authors would like to thank Ryoko Kataoka; EBMUD Operations staff (Joe Augustine, Trig Birkland, Robert Poole, Jay MacDonald, George Torres, Bill Byers, Frank Anderson) assisted with testing of the full-scale process; EBMUD Maintenance staff (Ike Bell, Jorge Garcia, Pat Virgin, Craig Patterson, Danny Devera, Todd Meyers, Doug Cooper, Terry Arnall, Mike Eleccion, Rick Treadwell, Bill Ledgewood, Kelvin Kirkeland) assisted with equipment installation and repairs; and Ken Gerstman, EBMUD Laboratory.
Donald Gray (Gabb), Paul Suto and Mark Chien are with the East Bay Municipal Utility District in Oakland, California. This article is based on a paper presented in 2007 at a Water Environment Federation conference.
REFERENCES
Metcalf and Eddy Inc. (2003) Wastewater Engineering, Treatment and Reuse, 4th ed.; McGraw-Hill: New York, NY.
Sanders, W.T.M., Veeken, A.H.M., Zeeman, G., and van Lier, J.B. (2003) “Analysis and optimization of the anaerobic digestion of the organic fraction of municipal solid waste.” In: Biomethanization of the organic fraction of municipal solid wastes. J. Mata-Alvarez, Editor. IWA Publishing.
Water Environment Federation (1998) Design of Municipal Wastewater Treatment Plants, 4th ed.; Manual of Practice 8; Alexandria, Virginia.
Sidebar p. 58
REGION 9 EPA ORGANICS RECYCLING INITIATIVES
UNDERSTANDING the relationship between materials management, or the lifecycle of a consumer product from raw material to disposal, and energy use is a critical component of any strategy to reduce greenhouse gas emissions. One way the U.S. Environmental Protection Agency (EPA) addresses materials management is through its Resource Conservation Challenge (RCC) – a national effort to conserve natural resources and energy by managing materials more efficiently. The program aims to reduce materials and energy usage at each stage of a material’s lifecycle through source reduction, reuse, diversion, recycling, composting and beneficial use of nonhazardous waste.
According to the 2006 Franklin Report, organic materials make up 26 percent of the waste being sent to landfills, making it the largest portion of the waste stream. The RCC aims to reduce the amount of organics being sent to landfills through recycling.
Organics recycling opportunities supported by EPA Region 9 are composting and anaerobic digestion. Composting, when managed properly, does not emit methane and also creates a valuable end product. On the other hand, food waste decomposition in a landfill results in significant uncaptured methane released to the atmosphere.
Anaerobic digestion of food waste at wastewater treatment facilities is a new concept that is very favorable in urban areas. Not only does excess capacity exist in many wastewater treatment facility digesters, but the addition of an energy-rich product, such as food waste, creates a much higher methane yield than biosolids alone. For example, research shows that the methane potential of cattle manure is around 25 m3 gas/ton, biosolids 120 m3 gas/ton and food waste anywhere from 220 to 650 m3 gas/ton depending on the richness of the waste. The end residual from a digester can be transported to a compost facility at a reduced volume and ultimately be used in the same beneficial way as compost.
The East Bay Municipal Utility District’s innovative waste diversion project (see main article) was funded through a U.S. EPA, Region 9 Resource Conservation Fund grant. In support of the RCC, Region 9 runs an annual Resource Conservation Fund grant solicitation that focuses on emerging issues in the waste industry, such as green building, construction and demolition debris, environmentally preferable purchasing, biogas, plastics recycling and organic materials management. For more information, visit www.epa.gov/ region09/funding/tribal-solid-waste. For a complete listing of all Federal Grants, see www.grants.gov. More details on the EPA Resource Conservation Challenge are available at http://www.epa.gov/rcc.
Cara Peck, USEPA Region 9