About five years ago, the Water Environment Federation (WEF) formally began using the term, water resource recovery facility (WRRF), in place of wastewater treatment plant. WEF noted that the name change reflects a changing paradigm in the water sector, focusing on recovery of water, nutrient and energy resources. In a press release announcing the change, WEF referenced codigestion as a resource recovery practice at WRRFs.
Increasingly, WRRFs in the U.S. with anaerobic digestion capacity are opening their gates to high strength organics such as commercial food waste and fats, oils and grease (FOG). “One driver in some regions is regulatory, where states are beginning to require generators over a certain size to divert food waste from landfill disposal,” notes Peter Loomis, a Technical Advisor with CDM Smith. “Another driver is economics. For a relatively small capital investment, WRRFs can get financial returns.”
The California Association of Sanitation Agencies estimates that existing infrastructure at government-owned WRRFs in California could accept up to 75 percent (7 million wet tons) of the food waste stream being landfilled in the state. The California Air Resources Board carried out a geospatial analysis indicating that food waste and wastewater treatment excess capacity are spatially correlated throughout California. The analysis found that all food waste currently landfilled could theoretically be consumed by wastewater treatment plants within 30 miles of these landfills.
“In our experience, whether a WRRF moves forward with codigestion is a top-down decision,” says Loomis. “Senior management needs to buy into the plan first, and then the operations and plant management will look at options to optimize the existing facility. In general, there is not a huge impact on the operations of the digester itself but more monitoring is needed.”
The primary capital investment related to codigestion is food waste and/or FOG receiving. Capability should be in place to remove heavy rocks, silverware and other items that come in with the loads. WRRFs receiving FOG will need screening, heated storage and transfer pumps, adds Loomis. “After food waste receiving, the next investment will be odor control in the receiving area and then gas handling. Food waste and FOG receiving facilities have historically demonstrated significant odors, potentially generating complaints. Odor control is needed on the front end, and more input to the digester will result in more gas production on the back end.”
Another consideration is the nutrient impact on the liquid stream that gets recycled back to the treatment plant after biosolids dewatering. “Generally speaking, there is a high ammonia load,” explains Loomis. “With biosolids, normally the ammonia load that returns to the liquids stream is in proportion to what is being processed. As a rule of thumb, there is one dry ton of solids for every one million gallons of sewage treated. Liquid separated from the solids — the centrate — is in proportion, nutrient wise, to the solids that are taken out. When WRRFs incorporate high strength organics and food waste, instead of one dry ton/million gallons of sewage, they are processing 1.25 to 1.5 dry tons for every million gallons, so there is more ammonia going back in the centrate.”
WRRFs typically have nutrient limits on the nitrogen and phosphorus that can be discharged, and the aeration basin’s ability to denitrify the ammonia can require more air to complete the process. This has the potential to increase costs on managing both the liquid and solids streams, he adds: “WRRFs can do a sidestream treatment of the centrate, but only a handful of facilities in the U.S. have installed that capability.”
East Bay Municipal Utility District
One WRRF examining the impact of excess nutrients from trucked in loads of high strength organics (HSOs) is the East Bay Municipal Utility District (EBMUD) in Oakland, California. EBMUD provides primary treatment for up to 320 million gallons/day (MGD), and secondary treatment for a maximum flow of 168 MGD. On average, about 63 million gallons of wastewater is treated every day.
EBMUD was a pioneer in codigestion of food waste, receiving its first trucked in load in May 2004 (see “Green Energy From Food Wastes At Wastewater Treatment Plant,” January 2008). EBMUD now receives about 4,000 truckloads per month for treatment and digestion, which enabled the utility to become the first WRRF in North America to produce more electricity onsite than is needed to run the treatment plant.
“We have about 250 customers, and service about 100 to 150 trucks per day,” notes Michael Hyatt, Associate Engineer, Resource Recovery, at EBMUD. “Two-thirds of the trucked in waste has low COD [chemical oxygen demand] and goes to the head of the plant. Of the one-third we take that is high strength, by volume it is 15 percent FOG, one percent food waste, although this is much more concentrated than the rest, 10 percent is animal protein waste and the remaining 74 to 75 percent are various types of process water, e.g, from the dairy and wine industries.”
Most of the HSOs arrive as a liquid. Food waste that is greater than 15 percent solids has to be slurried. All HSOs are loaded into two concrete cylinder blending tanks. The blending facility was built in 2014, along with a new receiving area. “We blend all the material going to the digesters to try to ensure a more uniform feed and avoid sudden variations,” explains Hyatt. All of the digesters receive the codigestion mix. This includes all of the trucked waste as well as the domestic wastewater treatment streams (primary sludge and waste activated sludge. He adds that in the earlier days of codigestion, “we would get a load of high strength material and put it into one digester, which could lead to an upset.”
The food waste is almost all from commercial generators, but EBMUD does get a small amount of preprocessed residential material that is delivered as a liquid. Generally food waste is around 30 percent solids, and in addition to slurrying, needs preprocessing to remove contamination before it is sent to the blend tanks.
EBMUD tests the first load of any new HSO being received, and sometimes inspects the waste stream generator’s production facility (e.g., if it is something less familiar to EBMUD). Constituents being measured include sulfur, salt and nitrogen. After testing the first loads, operators will spot check loads to track the content over time. “The dairy loads, for example, can be very salty,” says Hyatt. “Cheese waste has a high COD, which is good for the digester, but it does impact the level of salt in the centrate after dewatering.” Salt also is high in the trucked in brine waste (a low strength liquid) that is unloaded at the head of the plant.
Anaerobic digestion breaks down the proteins and the organics, but doesn’t remove sodium, phosphate or magnesium ions in the total dissolved solids (TDS), he adds. This can result in precipitation of some of the salts in the digestion tanks and on the pipe walls. EBMUD is evaluating the impact of the TDS levels on recycling of treated water. For example, cooling towers require low TDS whereas when recycled water is used for irrigation, the level can be higher. “We could put in a reverse osmosis membrane to remove the TDS but that is a very expensive option,” says Hyatt.
To manage hydrogen sulfide (H2S), EBMUD adds ferric chloride to the digester tanks, as the plant has limited ability to remove H2S in the gas phase. “With the growth of codigestion, we have had to monitor this much more closely and try to anticipate loads as the ferric addition takes some time to affect the process,” he notes.
Of the total nutrient load from wastewater treatment plants to the San Francisco Bay (where EBMUD’s treated water is discharged), EBMUD contributes approximately 20 percent, which is high relative to EBMUD’s flow, according to Hyatt. “This is due to nutrients in the trucked waste coming to the plant. The most effective place to add a treatment step would be on the centrate line, after digestion but before returning to the rest of the plant. Sidestream treatment on the centrate line is something we are actively researching.”