Onion Processor Upgrades Biogas To Power Fuel Cells

ONION waste, once a liability, is now an asset at Gill's Onions in Oxnard, California, where a $9.5 million project is converting onion waste into renewable energy and cattle feed. In 1983, Gill's started slicing and dicing onions for La Victoria Salsa. Over the years the company grew to become the country's largest producer of fresh-cut onions, processing almost one million pounds of onions and creating 300,000 pounds of waste each day.

Residuals from the process — the tops, tails and peels of the onions — were composted and spread on agricultural fields. "Land applying the waste worked until we got to a certain size and then it just overwhelmed the farming operations," says Steve Gill, co-owner of the business. Spreading the material on fields resulted in odor, runoff and pest problems. Costs associated with handling and disposing of the waste increased to $500,000 a year. "I had to find a solution to the problem," Gill adds.

The first alternative tried was making products out of the waste. "It did not work for cattle feed because cows did not like eating raw onions," recalls Bill Deaton, president of Deaton & Associates, LLC, and consultant on the project. Attempts to reduce the volume by grinding it up morphed into a study with the University of California, Davis, to see if Gill's could make biogas out of it, he explains.

Testing initially focused on digesting the solid waste streams. "Later they approached me and asked whether they could squeeze the juice from the solids and digest the juice," says Ruihong Zhang, professor at University of California, Davis. Bench-top tests digesting the solids and the juice found both waste streams to be highly digestible with good biogas yields.

Gill's opted to digest the juice. With space at the processing facility at a premium, a highrate anaerobic treatment system to process the juice required significantly less footprint and shorter retention times than a conventional solids digester. Experiments to extract the juice from the onion waste were successful with yields between 70 to 75 percent by weight, says Juan Josse with HDR, the process engineer and project manager who led the design of the facility. The remaining cake, which is 18 to 20 percent solids, was found to be good feed for cattle. "The farmers were willing to pay for it," he adds. "If we digested the entire peel and then dewatered the digestate we could not sell it as cattle feed. You have bacteria as part of the feed and this is not welcome as animal feed. Also most of the energy value is already extracted."

Gill's also discovered that it could extract polyphenol, an antioxidant for the nutraceuticals market, from the onion cake. The cake could still be sold as cattle feed after the extraction process.

WASTE PROCESSING

The juice extraction process uses two stages of size reduction and pressing to increase the yield. In the first stage, lime is added to the onion waste, which is ground in a maserator and then sent to a screw press to extract the juice. The lime increases juice yields by enhancing the traction between the waste and the outer screen in the screw press, preventing the onions from sliding horizontally against the screen, Josse explains. The highly alkaline lime also elevates the pH, which helps rupture the cell walls in the onion, releasing more juice.

The screw press produces onion juice with some suspended solids and onion cake. "We realized at this point that we had a juice yield of 50 percent, not the 75 percent we wanted," Josse says. A second stage is used to extract additional juice. The onion cake from the first press is sent to a second grinder for further size reduction. The finely cut cake is then fed to a smaller screw press that extracts the rest of the juice.

Conveyors move the remaining cake to a truck for transport to California farms for cattle feed. Juice collected from the two sets of presses is sent to a vibrator to separate out the small onion pieces suspended in the solution. The pieces are sent back to the second stage of the screw press for juice extraction. The process produces 30,000 gallons of juice and 20 tons of onion cake per day. The juice is high in soluble organics with about 60,000-ppm COD (chemical oxygen demand), Deaton says.

ANAEROBIC SLUDGE BED REACTOR

Onion juice is put into a 70,000-gallon acidification tank with mixers. After about two days in the tank, the juice is fed into the bottom of an anaerobic sludge bed (UASB) reactor supplied by Biothane in Camden, New Jersey.

An influent distribution system pulses the juice into the tank. The pulsing action serves to disperse the juice in the reactor and prevents tunneling of the influent through the thick sludge bed. Bacteria in the granular sludge bed start biodegrading the soluble organics in the juice and producing gas. The gas attaches to the granular sludge, causing some granules to float up through the bed and hydraulically forcing it through a series of baffle plates on a settler at the top of the tank where it is degasified. The degasified sludge granules sink back into the reactor while clear effluent flows over a weir near the top of the tank. Biogas collected below the reactor surface exits for treatment. The upflow velocity produced by the feed and effluent recirculation eliminates the need for mechanical mixing. Retention time for a UASB reactor typically averages 16 to 18 hours depending on the feedstock.

"We chose the Biothane reactor for the short residence time and robust performance," Deaton says. "They were also willing to work with us on our specification for not only COD removal but for biogas production." The square 110,000-gallon UASB reactor is painted black to absorb heat from the sun, helping to keep the tank warm. Waste heat captured from an 800-hp engine driving an air compressor used elsewhere in the facility maintains the reactor at mesophilic temperatures between 92° to 95°F. A brewery in St. Louis, Missouri provided 50,000 gallons of granular sludge that was 8 to 9- percent solids of seed the reactor. The microbe population adjusted quickly to its new diet of onion juice. Initially the granules grew too large but have since adapted to the new feedstock, Josse explains. "They are on fire, eating everything we give to them."

Biogas, produced at a rate of 100 to 110-cfm, contains about 70 percent methane. Effluent from the reactor is fed into an existing activated sludge plant at Gill's where it is aerated and further treated before being discharged into the City sewer.

FUEL CELLS

The biogas powers two 300-kW fuel cells generating 0.6 MW of electricity to satisfy 75 percent of Gill's base load power requirements. The fuel cells, supplied by Fuel Cell Energy in Danbury, Connecticut, operate on biogas and natural gas.

Fuel cells were selected over other technologies for their efficiency, exempt status for air quality permitting and available incentives. "Fuel cells are 47 percent efficient versus internal combustion (IC) engines that are 35 percent efficient," Gill says. Utilizing waste heat from the fuel cells, which will occur in the next phase of the project, will push the overall efficiency of the system to 90 percent."The fuel cell had the advantage that we did not have to obtain an air permit," Deaton says. "It was exempt from air pollution control testing." Strict air quality regulations already limit the number and operation of stationary IC engines in Ventura County, where Gill's Onions is located.

The project received $2.7 million in self-generation incentive funds from Sempra Energy, the local utility, and a 30 percent federal investment tax credit for installing the renewable energy process. Sempra's incentive program provides businesses with financial incentives to install electricity generation technologies and pays higher incentives for technologies that run on renewable fuels. The utility current pays $4,500 per kW of installed capacity for fuel cells running on renewable power.

BIOGAS CLEANUP

One challenge the project faced was cleaning the biogas to meet fuel cell specifications. Fuel cells are extremely sensitive to sulfur, with levels limited to 100 parts per billion (0.1 ppm) or less, explains Ted Barnes, principal engineer at the Gas Technology Institute (GTI) in Des Plaines, Illinois. This is significantly cleaner than biomethane quality standards for pipeline injection.

Onions are high in sulfur. It's the sulfates in the onions that make you cry and also why they smell and taste so good, Barnes explains. These sulfur compounds end up in the juice and during anaerobic digestion are volatized into biogas creating high concentrations — more than 5,000 ppm of hydrogen sulfide (H2S) and organic sulfur compounds. When the project was under development, commercial manufacturers of biogas cleanup technologies could not guarantee their systems to meet sulfur removal specifications. "We had to develop new technology to remove the sulfur," Deaton says. Gill's project team hooked up with the GTI to research development of a gas purification system. Funding for the research was provided through a $500,000 grant from the California Energy Commission.

Research focused on testing and developing media to capture both the inorganic and organic sulfur compounds. "We found several media that worked," Deaton says. "We had to design our system so that those medias were in the proportion of the gas impurities. We also went to a lot of effort to understand what pressure, temperature and humidity conditions worked best for the separation and optimize all the variables."

The first step in the process devised to clean up the biogas uses a two-stage iron sponge to get H2S levels down below 1 ppm, Barnes says. A compression and drying system employs a glycol chiller to remove moisture from the gas. Recovered heat from the compressor reheats the gas. The final step passes the gas through two vessels in series, each with different media, to remove the organic sulfurs that made it through the iron sponge. The two media in combination proved very effective at removing the organic sulfur compounds. Final polishing of the gas uses filters to remove particulates carried over from the prior cleanup stage. When the iron sponge media becomes saturated with sulfur, it can be regenerated two to three times. "You put oxygen and air through it in the presence of water," explains Barnes. "The sulfur comes out in elemental sulfur form." The second polishing media can only be regenerated off-site. When this media is spent it needs to be replaced. Unlike pipeline injection, CO2 does not need to be removed from the biogas. The molten carbonated fuel cells used on the project tolerate CO2.

To date, the gas management system is working well, Gill says. Sampling is currently used to monitor gas quality. Installation of sensors is underway to automate the process.

VARIABLE GAS FLOW

Gas generation in a UASB reactor fluctuates, with the gas flows going up and down in 6 to 7 minute cycles, says Josse. Bacteria in the granular bed produce tiny biogas bubbles as they consume the organic compounds. The bubbles combine to form larger bubbles that are finally able to break through the sludge bed. The physics of forming the larger bubbles has a "certain periodicity to it."

The team needed to devise a process to overcome the variable gas flow from the digester, a situation that had resulted in inefficiencies in other fuel cell applications, Josse adds. The fuel cells, which operate on both natural gas and biogas, are setup to give preference to biogas. If the devices are not getting enough biogas, they need to compensate with natural gas. Variations in the flow of the biogas are too abrupt for the fuel cells to respond appropriately. As a result, the cells ends up using more natural gas and flaring biogas, which increases operating costs and wastes energy, he says.

To solve the problem, a gas holder with a double membrane and an ultrasonic sensor was devised to absorb the fluctuations. If the level of the gas starts to drop, the sensor sends an advanced signal to the fuel cells to slightly close the valve supplying biogas and slightly open the natural gas valve. "It works wonderfully," Josse says. "We do not run the flare. Basically everything gets used and there is very little natural gas supplementation."

PROJECT ECONOMICS

Using biogas in the fuel cells is saving $50,000 to $60,000 per month in electricity purchases, Gill says. Another $500,000 is saved annually by eliminating the hauling and spreading of onion waste in the fields. The project is also earning money by selling the onion cake as cattle feed. Taking into account the savings, sales of by-products, selfgeneration credits and investment tax credits, the $9.5 million project expects payback in five years.

Everyone associated with the renewable energy project emphasizes the level of cooperation and teamwork. "We were successful in pulling together all the right resources, getting the funding and managing the overall project during an awkward time in our country," Deaton says. "It is really a boon for Gill's Onions. They are taking a problem and making it into a profit center."

Diane Greer is a Contributing Editor to BioCycle. Bill Deaton will be speaking at the 25th Annual BioCycle West Coast Conference on the Gill's Onion project